pcircle - A Suite of Scalable Parallel File System Tools

DOE Office of Scientific and Technical Information (OSTI.GOV)

WANG, FEIYI

2015-10-01

Most of the software related to file system are written for conventional local file system, they are serialized and can't take advantage of the benefit of a large scale parallel file system. "pcircle" software builds on top of ubiquitous MPI in cluster computing environment and "work-stealing" pattern to provide a scalable, high-performance suite of file system tools. In particular - it implemented parallel data copy and parallel data checksumming, with advanced features such as async progress report, checkpoint and restart, as well as integrity checking.

Parallel multigrid smoothing: polynomial versus Gauss-Seidel

NASA Astrophysics Data System (ADS)

Adams, Mark; Brezina, Marian; Hu, Jonathan; Tuminaro, Ray

2003-07-01

Gauss-Seidel is often the smoother of choice within multigrid applications. In the context of unstructured meshes, however, maintaining good parallel efficiency is difficult with multiplicative iterative methods such as Gauss-Seidel. This leads us to consider alternative smoothers. We discuss the computational advantages of polynomial smoothers within parallel multigrid algorithms for positive definite symmetric systems. Two particular polynomials are considered: Chebyshev and a multilevel specific polynomial. The advantages of polynomial smoothing over traditional smoothers such as Gauss-Seidel are illustrated on several applications: Poisson's equation, thin-body elasticity, and eddy current approximations to Maxwell's equations. While parallelizing the Gauss-Seidel method typically involves a compromise between a scalable convergence rate and maintaining high flop rates, polynomial smoothers achieve parallel scalable multigrid convergence rates without sacrificing flop rates. We show that, although parallel computers are the main motivation, polynomial smoothers are often surprisingly competitive with Gauss-Seidel smoothers on serial machines.

Scalable Domain Decomposed Monte Carlo Particle Transport

NASA Astrophysics Data System (ADS)

O'Brien, Matthew Joseph

In this dissertation, we present the parallel algorithms necessary to run domain decomposed Monte Carlo particle transport on large numbers of processors (millions of processors). Previous algorithms were not scalable, and the parallel overhead became more computationally costly than the numerical simulation. The main algorithms we consider are: • Domain decomposition of constructive solid geometry: enables extremely large calculations in which the background geometry is too large to fit in the memory of a single computational node. • Load Balancing: keeps the workload per processor as even as possible so the calculation runs efficiently. • Global Particle Find: if particles are on the wrong processor, globally resolve their locations to the correct processor based on particle coordinate and background domain. • Visualizing constructive solid geometry, sourcing particles, deciding that particle streaming communication is completed and spatial redecomposition. These algorithms are some of the most important parallel algorithms required for domain decomposed Monte Carlo particle transport. We demonstrate that our previous algorithms were not scalable, prove that our new algorithms are scalable, and run some of the algorithms up to 2 million MPI processes on the Sequoia supercomputer.

Scalable isosurface visualization of massive datasets on commodity off-the-shelf clusters

PubMed Central

Bajaj, Chandrajit

2009-01-01

Tomographic imaging and computer simulations are increasingly yielding massive datasets. Interactive and exploratory visualizations have rapidly become indispensable tools to study large volumetric imaging and simulation data. Our scalable isosurface visualization framework on commodity off-the-shelf clusters is an end-to-end parallel and progressive platform, from initial data access to the final display. Interactive browsing of extracted isosurfaces is made possible by using parallel isosurface extraction, and rendering in conjunction with a new specialized piece of image compositing hardware called Metabuffer. In this paper, we focus on the back end scalability by introducing a fully parallel and out-of-core isosurface extraction algorithm. It achieves scalability by using both parallel and out-of-core processing and parallel disks. It statically partitions the volume data to parallel disks with a balanced workload spectrum, and builds I/O-optimal external interval trees to minimize the number of I/O operations of loading large data from disk. We also describe an isosurface compression scheme that is efficient for progress extraction, transmission and storage of isosurfaces. PMID:19756231

Performance and Scalability of the NAS Parallel Benchmarks in Java

NASA Technical Reports Server (NTRS)

Frumkin, Michael A.; Schultz, Matthew; Jin, Haoqiang; Yan, Jerry; Biegel, Bryan A. (Technical Monitor)

2002-01-01

Several features make Java an attractive choice for scientific applications. In order to gauge the applicability of Java to Computational Fluid Dynamics (CFD), we have implemented the NAS (NASA Advanced Supercomputing) Parallel Benchmarks in Java. The performance and scalability of the benchmarks point out the areas where improvement in Java compiler technology and in Java thread implementation would position Java closer to Fortran in the competition for scientific applications.

A Systems Approach to Scalable Transportation Network Modeling

DOE Office of Scientific and Technical Information (OSTI.GOV)

Perumalla, Kalyan S

2006-01-01

Emerging needs in transportation network modeling and simulation are raising new challenges with respect to scal-ability of network size and vehicular traffic intensity, speed of simulation for simulation-based optimization, and fidel-ity of vehicular behavior for accurate capture of event phe-nomena. Parallel execution is warranted to sustain the re-quired detail, size and speed. However, few parallel simulators exist for such applications, partly due to the challenges underlying their development. Moreover, many simulators are based on time-stepped models, which can be computationally inefficient for the purposes of modeling evacuation traffic. Here an approach is presented to de-signing a simulator with memory andmore » speed efficiency as the goals from the outset, and, specifically, scalability via parallel execution. The design makes use of discrete event modeling techniques as well as parallel simulation meth-ods. Our simulator, called SCATTER, is being developed, incorporating such design considerations. Preliminary per-formance results are presented on benchmark road net-works, showing scalability to one million vehicles simu-lated on one processor.« less

Application of a Scalable, Parallel, Unstructured-Grid-Based Navier-Stokes Solver

NASA Technical Reports Server (NTRS)

Parikh, Paresh

2001-01-01

A parallel version of an unstructured-grid based Navier-Stokes solver, USM3Dns, previously developed for efficient operation on a variety of parallel computers, has been enhanced to incorporate upgrades made to the serial version. The resultant parallel code has been extensively tested on a variety of problems of aerospace interest and on two sets of parallel computers to understand and document its characteristics. An innovative grid renumbering construct and use of non-blocking communication are shown to produce superlinear computing performance. Preliminary results from parallelization of a recently introduced "porous surface" boundary condition are also presented.

Proxy-equation paradigm: A strategy for massively parallel asynchronous computations

NASA Astrophysics Data System (ADS)

Mittal, Ankita; Girimaji, Sharath

2017-09-01

Massively parallel simulations of transport equation systems call for a paradigm change in algorithm development to achieve efficient scalability. Traditional approaches require time synchronization of processing elements (PEs), which severely restricts scalability. Relaxing synchronization requirement introduces error and slows down convergence. In this paper, we propose and develop a novel "proxy equation" concept for a general transport equation that (i) tolerates asynchrony with minimal added error, (ii) preserves convergence order and thus, (iii) expected to scale efficiently on massively parallel machines. The central idea is to modify a priori the transport equation at the PE boundaries to offset asynchrony errors. Proof-of-concept computations are performed using a one-dimensional advection (convection) diffusion equation. The results demonstrate the promise and advantages of the present strategy.

A Multi-Time Scale Morphable Software Milieu for Polymorphous Computing Architectures (PCA) - Composable, Scalable Systems

DTIC Science & Technology

2004-10-01

MONITORING AGENCY NAME(S) AND ADDRESS(ES) Defense Advanced Research Projects Agency AFRL/IFTC 3701 North Fairfax Drive...Scalable Parallel Libraries for Large-Scale Concurrent Applications," Technical Report UCRL -JC-109251, Lawrence Livermore National Laboratory

Parallel peak pruning for scalable SMP contour tree computation

DOE Office of Scientific and Technical Information (OSTI.GOV)

Carr, Hamish A.; Weber, Gunther H.; Sewell, Christopher M.

As data sets grow to exascale, automated data analysis and visualisation are increasingly important, to intermediate human understanding and to reduce demands on disk storage via in situ analysis. Trends in architecture of high performance computing systems necessitate analysis algorithms to make effective use of combinations of massively multicore and distributed systems. One of the principal analytic tools is the contour tree, which analyses relationships between contours to identify features of more than local importance. Unfortunately, the predominant algorithms for computing the contour tree are explicitly serial, and founded on serial metaphors, which has limited the scalability of this formmore » of analysis. While there is some work on distributed contour tree computation, and separately on hybrid GPU-CPU computation, there is no efficient algorithm with strong formal guarantees on performance allied with fast practical performance. Here in this paper, we report the first shared SMP algorithm for fully parallel contour tree computation, withfor-mal guarantees of O(lgnlgt) parallel steps and O(n lgn) work, and implementations with up to 10x parallel speed up in OpenMP and up to 50x speed up in NVIDIA Thrust.« less

Novel Scalable 3-D MT Inverse Solver

NASA Astrophysics Data System (ADS)

Kuvshinov, A. V.; Kruglyakov, M.; Geraskin, A.

2016-12-01

We present a new, robust and fast, three-dimensional (3-D) magnetotelluric (MT) inverse solver. As a forward modelling engine a highly-scalable solver extrEMe [1] is used. The (regularized) inversion is based on an iterative gradient-type optimization (quasi-Newton method) and exploits adjoint sources approach for fast calculation of the gradient of the misfit. The inverse solver is able to deal with highly detailed and contrasting models, allows for working (separately or jointly) with any type of MT (single-site and/or inter-site) responses, and supports massive parallelization. Different parallelization strategies implemented in the code allow for optimal usage of available computational resources for a given problem set up. To parameterize an inverse domain a mask approach is implemented, which means that one can merge any subset of forward modelling cells in order to account for (usually) irregular distribution of observation sites. We report results of 3-D numerical experiments aimed at analysing the robustness, performance and scalability of the code. In particular, our computational experiments carried out at different platforms ranging from modern laptops to high-performance clusters demonstrate practically linear scalability of the code up to thousands of nodes. 1. Kruglyakov, M., A. Geraskin, A. Kuvshinov, 2016. Novel accurate and scalable 3-D MT forward solver based on a contracting integral equation method, Computers and Geosciences, in press.

Performance prediction: A case study using a multi-ring KSR-1 machine

NASA Technical Reports Server (NTRS)

Sun, Xian-He; Zhu, Jianping

1995-01-01

While computers with tens of thousands of processors have successfully delivered high performance power for solving some of the so-called 'grand-challenge' applications, the notion of scalability is becoming an important metric in the evaluation of parallel machine architectures and algorithms. In this study, the prediction of scalability and its application are carefully investigated. A simple formula is presented to show the relation between scalability, single processor computing power, and degradation of parallelism. A case study is conducted on a multi-ring KSR1 shared virtual memory machine. Experimental and theoretical results show that the influence of topology variation of an architecture is predictable. Therefore, the performance of an algorithm on a sophisticated, heirarchical architecture can be predicted and the best algorithm-machine combination can be selected for a given application.

Large-scale parallel genome assembler over cloud computing environment.

PubMed

Das, Arghya Kusum; Koppa, Praveen Kumar; Goswami, Sayan; Platania, Richard; Park, Seung-Jong

2017-06-01

The size of high throughput DNA sequencing data has already reached the terabyte scale. To manage this huge volume of data, many downstream sequencing applications started using locality-based computing over different cloud infrastructures to take advantage of elastic (pay as you go) resources at a lower cost. However, the locality-based programming model (e.g. MapReduce) is relatively new. Consequently, developing scalable data-intensive bioinformatics applications using this model and understanding the hardware environment that these applications require for good performance, both require further research. In this paper, we present a de Bruijn graph oriented Parallel Giraph-based Genome Assembler (GiGA), as well as the hardware platform required for its optimal performance. GiGA uses the power of Hadoop (MapReduce) and Giraph (large-scale graph analysis) to achieve high scalability over hundreds of compute nodes by collocating the computation and data. GiGA achieves significantly higher scalability with competitive assembly quality compared to contemporary parallel assemblers (e.g. ABySS and Contrail) over traditional HPC cluster. Moreover, we show that the performance of GiGA is significantly improved by using an SSD-based private cloud infrastructure over traditional HPC cluster. We observe that the performance of GiGA on 256 cores of this SSD-based cloud infrastructure closely matches that of 512 cores of traditional HPC cluster.

A comprehensive study of MPI parallelism in three-dimensional discrete element method (DEM) simulation of complex-shaped granular particles

NASA Astrophysics Data System (ADS)

Yan, Beichuan; Regueiro, Richard A.

2018-02-01

A three-dimensional (3D) DEM code for simulating complex-shaped granular particles is parallelized using message-passing interface (MPI). The concepts of link-block, ghost/border layer, and migration layer are put forward for design of the parallel algorithm, and theoretical scalability function of 3-D DEM scalability and memory usage is derived. Many performance-critical implementation details are managed optimally to achieve high performance and scalability, such as: minimizing communication overhead, maintaining dynamic load balance, handling particle migrations across block borders, transmitting C++ dynamic objects of particles between MPI processes efficiently, eliminating redundant contact information between adjacent MPI processes. The code executes on multiple US Department of Defense (DoD) supercomputers and tests up to 2048 compute nodes for simulating 10 million three-axis ellipsoidal particles. Performance analyses of the code including speedup, efficiency, scalability, and granularity across five orders of magnitude of simulation scale (number of particles) are provided, and they demonstrate high speedup and excellent scalability. It is also discovered that communication time is a decreasing function of the number of compute nodes in strong scaling measurements. The code's capability of simulating a large number of complex-shaped particles on modern supercomputers will be of value in both laboratory studies on micromechanical properties of granular materials and many realistic engineering applications involving granular materials.

Parallel scalability of Hartree-Fock calculations

NASA Astrophysics Data System (ADS)

Chow, Edmond; Liu, Xing; Smelyanskiy, Mikhail; Hammond, Jeff R.

2015-03-01

Quantum chemistry is increasingly performed using large cluster computers consisting of multiple interconnected nodes. For a fixed molecular problem, the efficiency of a calculation usually decreases as more nodes are used, due to the cost of communication between the nodes. This paper empirically investigates the parallel scalability of Hartree-Fock calculations. The construction of the Fock matrix and the density matrix calculation are analyzed separately. For the former, we use a parallelization of Fock matrix construction based on a static partitioning of work followed by a work stealing phase. For the latter, we use density matrix purification from the linear scaling methods literature, but without using sparsity. When using large numbers of nodes for moderately sized problems, density matrix computations are network-bandwidth bound, making purification methods potentially faster than eigendecomposition methods.

Development and Applications of a Modular Parallel Process for Large Scale Fluid/Structures Problems

NASA Technical Reports Server (NTRS)

Guruswamy, Guru P.; Byun, Chansup; Kwak, Dochan (Technical Monitor)

2001-01-01

A modular process that can efficiently solve large scale multidisciplinary problems using massively parallel super computers is presented. The process integrates disciplines with diverse physical characteristics by retaining the efficiency of individual disciplines. Computational domain independence of individual disciplines is maintained using a meta programming approach. The process integrates disciplines without affecting the combined performance. Results are demonstrated for large scale aerospace problems on several supercomputers. The super scalability and portability of the approach is demonstrated on several parallel computers.

The Open Connectome Project Data Cluster: Scalable Analysis and Vision for High-Throughput Neuroscience.

PubMed

Burns, Randal; Roncal, William Gray; Kleissas, Dean; Lillaney, Kunal; Manavalan, Priya; Perlman, Eric; Berger, Daniel R; Bock, Davi D; Chung, Kwanghun; Grosenick, Logan; Kasthuri, Narayanan; Weiler, Nicholas C; Deisseroth, Karl; Kazhdan, Michael; Lichtman, Jeff; Reid, R Clay; Smith, Stephen J; Szalay, Alexander S; Vogelstein, Joshua T; Vogelstein, R Jacob

2013-01-01

We describe a scalable database cluster for the spatial analysis and annotation of high-throughput brain imaging data, initially for 3-d electron microscopy image stacks, but for time-series and multi-channel data as well. The system was designed primarily for workloads that build connectomes - neural connectivity maps of the brain-using the parallel execution of computer vision algorithms on high-performance compute clusters. These services and open-science data sets are publicly available at openconnecto.me. The system design inherits much from NoSQL scale-out and data-intensive computing architectures. We distribute data to cluster nodes by partitioning a spatial index. We direct I/O to different systems-reads to parallel disk arrays and writes to solid-state storage-to avoid I/O interference and maximize throughput. All programming interfaces are RESTful Web services, which are simple and stateless, improving scalability and usability. We include a performance evaluation of the production system, highlighting the effec-tiveness of spatial data organization.

The Open Connectome Project Data Cluster: Scalable Analysis and Vision for High-Throughput Neuroscience

PubMed Central

Burns, Randal; Roncal, William Gray; Kleissas, Dean; Lillaney, Kunal; Manavalan, Priya; Perlman, Eric; Berger, Daniel R.; Bock, Davi D.; Chung, Kwanghun; Grosenick, Logan; Kasthuri, Narayanan; Weiler, Nicholas C.; Deisseroth, Karl; Kazhdan, Michael; Lichtman, Jeff; Reid, R. Clay; Smith, Stephen J.; Szalay, Alexander S.; Vogelstein, Joshua T.; Vogelstein, R. Jacob

2013-01-01

We describe a scalable database cluster for the spatial analysis and annotation of high-throughput brain imaging data, initially for 3-d electron microscopy image stacks, but for time-series and multi-channel data as well. The system was designed primarily for workloads that build connectomes— neural connectivity maps of the brain—using the parallel execution of computer vision algorithms on high-performance compute clusters. These services and open-science data sets are publicly available at openconnecto.me. The system design inherits much from NoSQL scale-out and data-intensive computing architectures. We distribute data to cluster nodes by partitioning a spatial index. We direct I/O to different systems—reads to parallel disk arrays and writes to solid-state storage—to avoid I/O interference and maximize throughput. All programming interfaces are RESTful Web services, which are simple and stateless, improving scalability and usability. We include a performance evaluation of the production system, highlighting the effec-tiveness of spatial data organization. PMID:24401992

Long-range interactions and parallel scalability in molecular simulations

NASA Astrophysics Data System (ADS)

Patra, Michael; Hyvönen, Marja T.; Falck, Emma; Sabouri-Ghomi, Mohsen; Vattulainen, Ilpo; Karttunen, Mikko

2007-01-01

Typical biomolecular systems such as cellular membranes, DNA, and protein complexes are highly charged. Thus, efficient and accurate treatment of electrostatic interactions is of great importance in computational modeling of such systems. We have employed the GROMACS simulation package to perform extensive benchmarking of different commonly used electrostatic schemes on a range of computer architectures (Pentium-4, IBM Power 4, and Apple/IBM G5) for single processor and parallel performance up to 8 nodes—we have also tested the scalability on four different networks, namely Infiniband, GigaBit Ethernet, Fast Ethernet, and nearly uniform memory architecture, i.e. communication between CPUs is possible by directly reading from or writing to other CPUs' local memory. It turns out that the particle-mesh Ewald method (PME) performs surprisingly well and offers competitive performance unless parallel runs on PC hardware with older network infrastructure are needed. Lipid bilayers of sizes 128, 512 and 2048 lipid molecules were used as the test systems representing typical cases encountered in biomolecular simulations. Our results enable an accurate prediction of computational speed on most current computing systems, both for serial and parallel runs. These results should be helpful in, for example, choosing the most suitable configuration for a small departmental computer cluster.

Execution of a parallel edge-based Navier-Stokes solver on commodity graphics processor units

NASA Astrophysics Data System (ADS)

Corral, Roque; Gisbert, Fernando; Pueblas, Jesus

2017-02-01

The implementation of an edge-based three-dimensional Reynolds Average Navier-Stokes solver for unstructured grids able to run on multiple graphics processing units (GPUs) is presented. Loops over edges, which are the most time-consuming part of the solver, have been written to exploit the massively parallel capabilities of GPUs. Non-blocking communications between parallel processes and between the GPU and the central processor unit (CPU) have been used to enhance code scalability. The code is written using a mixture of C++ and OpenCL, to allow the execution of the source code on GPUs. The Message Passage Interface (MPI) library is used to allow the parallel execution of the solver on multiple GPUs. A comparative study of the solver parallel performance is carried out using a cluster of CPUs and another of GPUs. It is shown that a single GPU is up to 64 times faster than a single CPU core. The parallel scalability of the solver is mainly degraded due to the loss of computing efficiency of the GPU when the size of the case decreases. However, for large enough grid sizes, the scalability is strongly improved. A cluster featuring commodity GPUs and a high bandwidth network is ten times less costly and consumes 33% less energy than a CPU-based cluster with an equivalent computational power.

Parallel Algorithms for the Exascale Era

DOE Office of Scientific and Technical Information (OSTI.GOV)

Robey, Robert W.

New parallel algorithms are needed to reach the Exascale level of parallelism with millions of cores. We look at some of the research developed by students in projects at LANL. The research blends ideas from the early days of computing while weaving in the fresh approach brought by students new to the field of high performance computing. We look at reproducibility of global sums and why it is important to parallel computing. Next we look at how the concept of hashing has led to the development of more scalable algorithms suitable for next-generation parallel computers. Nearly all of this workmore » has been done by undergraduates and published in leading scientific journals.« less

A Scalable O(N) Algorithm for Large-Scale Parallel First-Principles Molecular Dynamics Simulations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Osei-Kuffuor, Daniel; Fattebert, Jean-Luc

2014-01-01

Traditional algorithms for first-principles molecular dynamics (FPMD) simulations only gain a modest capability increase from current petascale computers, due to their O(N 3) complexity and their heavy use of global communications. To address this issue, we are developing a truly scalable O(N) complexity FPMD algorithm, based on density functional theory (DFT), which avoids global communications. The computational model uses a general nonorthogonal orbital formulation for the DFT energy functional, which requires knowledge of selected elements of the inverse of the associated overlap matrix. We present a scalable algorithm for approximately computing selected entries of the inverse of the overlap matrix,more » based on an approximate inverse technique, by inverting local blocks corresponding to principal submatrices of the global overlap matrix. The new FPMD algorithm exploits sparsity and uses nearest neighbor communication to provide a computational scheme capable of extreme scalability. Accuracy is controlled by the mesh spacing of the finite difference discretization, the size of the localization regions in which the electronic orbitals are confined, and a cutoff beyond which the entries of the overlap matrix can be omitted when computing selected entries of its inverse. We demonstrate the algorithm's excellent parallel scaling for up to O(100K) atoms on O(100K) processors, with a wall-clock time of O(1) minute per molecular dynamics time step.« less

A derivation and scalable implementation of the synchronous parallel kinetic Monte Carlo method for simulating long-time dynamics

NASA Astrophysics Data System (ADS)

Byun, Hye Suk; El-Naggar, Mohamed Y.; Kalia, Rajiv K.; Nakano, Aiichiro; Vashishta, Priya

2017-10-01

Kinetic Monte Carlo (KMC) simulations are used to study long-time dynamics of a wide variety of systems. Unfortunately, the conventional KMC algorithm is not scalable to larger systems, since its time scale is inversely proportional to the simulated system size. A promising approach to resolving this issue is the synchronous parallel KMC (SPKMC) algorithm, which makes the time scale size-independent. This paper introduces a formal derivation of the SPKMC algorithm based on local transition-state and time-dependent Hartree approximations, as well as its scalable parallel implementation based on a dual linked-list cell method. The resulting algorithm has achieved a weak-scaling parallel efficiency of 0.935 on 1024 Intel Xeon processors for simulating biological electron transfer dynamics in a 4.2 billion-heme system, as well as decent strong-scaling parallel efficiency. The parallel code has been used to simulate a lattice of cytochrome complexes on a bacterial-membrane nanowire, and it is broadly applicable to other problems such as computational synthesis of new materials.

Speeding up parallel processing

NASA Technical Reports Server (NTRS)

Denning, Peter J.

1988-01-01

In 1967 Amdahl expressed doubts about the ultimate utility of multiprocessors. The formulation, now called Amdahl's law, became part of the computing folklore and has inspired much skepticism about the ability of the current generation of massively parallel processors to efficiently deliver all their computing power to programs. The widely publicized recent results of a group at Sandia National Laboratory, which showed speedup on a 1024 node hypercube of over 500 for three fixed size problems and over 1000 for three scalable problems, have convincingly challenged this bit of folklore and have given new impetus to parallel scientific computing.

A comparative study of serial and parallel aeroelastic computations of wings

NASA Technical Reports Server (NTRS)

Byun, Chansup; Guruswamy, Guru P.

1994-01-01

A procedure for computing the aeroelasticity of wings on parallel multiple-instruction, multiple-data (MIMD) computers is presented. In this procedure, fluids are modeled using Euler equations, and structures are modeled using modal or finite element equations. The procedure is designed in such a way that each discipline can be developed and maintained independently by using a domain decomposition approach. In the present parallel procedure, each computational domain is scalable. A parallel integration scheme is used to compute aeroelastic responses by solving fluid and structural equations concurrently. The computational efficiency issues of parallel integration of both fluid and structural equations are investigated in detail. This approach, which reduces the total computational time by a factor of almost 2, is demonstrated for a typical aeroelastic wing by using various numbers of processors on the Intel iPSC/860.

Scalability of a Low-Cost Multi-Teraflop Linux Cluster for High-End Classical Atomistic and Quantum Mechanical Simulations

NASA Technical Reports Server (NTRS)

Kikuchi, Hideaki; Kalia, Rajiv K.; Nakano, Aiichiro; Vashishta, Priya; Shimojo, Fuyuki; Saini, Subhash

2003-01-01

Scalability of a low-cost, Intel Xeon-based, multi-Teraflop Linux cluster is tested for two high-end scientific applications: Classical atomistic simulation based on the molecular dynamics method and quantum mechanical calculation based on the density functional theory. These scalable parallel applications use space-time multiresolution algorithms and feature computational-space decomposition, wavelet-based adaptive load balancing, and spacefilling-curve-based data compression for scalable I/O. Comparative performance tests are performed on a 1,024-processor Linux cluster and a conventional higher-end parallel supercomputer, 1,184-processor IBM SP4. The results show that the performance of the Linux cluster is comparable to that of the SP4. We also study various effects, such as the sharing of memory and L2 cache among processors, on the performance.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Bennett, Janine Camille; Thompson, David; Pebay, Philippe Pierre

Statistical analysis is typically used to reduce the dimensionality of and infer meaning from data. A key challenge of any statistical analysis package aimed at large-scale, distributed data is to address the orthogonal issues of parallel scalability and numerical stability. Many statistical techniques, e.g., descriptive statistics or principal component analysis, are based on moments and co-moments and, using robust online update formulas, can be computed in an embarrassingly parallel manner, amenable to a map-reduce style implementation. In this paper we focus on contingency tables, through which numerous derived statistics such as joint and marginal probability, point-wise mutual information, information entropy,more » and {chi}{sup 2} independence statistics can be directly obtained. However, contingency tables can become large as data size increases, requiring a correspondingly large amount of communication between processors. This potential increase in communication prevents optimal parallel speedup and is the main difference with moment-based statistics (which we discussed in [1]) where the amount of inter-processor communication is independent of data size. Here we present the design trade-offs which we made to implement the computation of contingency tables in parallel. We also study the parallel speedup and scalability properties of our open source implementation. In particular, we observe optimal speed-up and scalability when the contingency statistics are used in their appropriate context, namely, when the data input is not quasi-diffuse.« less

A Parallel Ghosting Algorithm for The Flexible Distributed Mesh Database

DOE PAGES

Mubarak, Misbah; Seol, Seegyoung; Lu, Qiukai; ...

2013-01-01

Critical to the scalability of parallel adaptive simulations are parallel control functions including load balancing, reduced inter-process communication and optimal data decomposition. In distributed meshes, many mesh-based applications frequently access neighborhood information for computational purposes which must be transmitted efficiently to avoid parallel performance degradation when the neighbors are on different processors. This article presents a parallel algorithm of creating and deleting data copies, referred to as ghost copies, which localize neighborhood data for computation purposes while minimizing inter-process communication. The key characteristics of the algorithm are: (1) It can create ghost copies of any permissible topological order in amore » 1D, 2D or 3D mesh based on selected adjacencies. (2) It exploits neighborhood communication patterns during the ghost creation process thus eliminating all-to-all communication. (3) For applications that need neighbors of neighbors, the algorithm can create n number of ghost layers up to a point where the whole partitioned mesh can be ghosted. Strong and weak scaling results are presented for the IBM BG/P and Cray XE6 architectures up to a core count of 32,768 processors. The algorithm also leads to scalable results when used in a parallel super-convergent patch recovery error estimator, an application that frequently accesses neighborhood data to carry out computation.« less

Scalable Parallel Density-based Clustering and Applications

NASA Astrophysics Data System (ADS)

Patwary, Mostofa Ali

2014-04-01

Recently, density-based clustering algorithms (DBSCAN and OPTICS) have gotten significant attention of the scientific community due to their unique capability of discovering arbitrary shaped clusters and eliminating noise data. These algorithms have several applications, which require high performance computing, including finding halos and subhalos (clusters) from massive cosmology data in astrophysics, analyzing satellite images, X-ray crystallography, and anomaly detection. However, parallelization of these algorithms are extremely challenging as they exhibit inherent sequential data access order, unbalanced workload resulting in low parallel efficiency. To break the data access sequentiality and to achieve high parallelism, we develop new parallel algorithms, both for DBSCAN and OPTICS, designed using graph algorithmic techniques. For example, our parallel DBSCAN algorithm exploits the similarities between DBSCAN and computing connected components. Using datasets containing up to a billion floating point numbers, we show that our parallel density-based clustering algorithms significantly outperform the existing algorithms, achieving speedups up to 27.5 on 40 cores on shared memory architecture and speedups up to 5,765 using 8,192 cores on distributed memory architecture. In our experiments, we found that while achieving the scalability, our algorithms produce clustering results with comparable quality to the classical algorithms.

A Massively Parallel Code for Polarization Calculations

NASA Astrophysics Data System (ADS)

Akiyama, Shizuka; Höflich, Peter

2001-03-01

We present an implementation of our Monte-Carlo radiation transport method for rapidly expanding, NLTE atmospheres for massively parallel computers which utilizes both the distributed and shared memory models. This allows us to take full advantage of the fast communication and low latency inherent to nodes with multiple CPUs, and to stretch the limits of scalability with the number of nodes compared to a version which is based on the shared memory model. Test calculations on a local 20-node Beowulf cluster with dual CPUs showed an improved scalability by about 40%.

Analysis of scalability of high-performance 3D image processing platform for virtual colonoscopy

NASA Astrophysics Data System (ADS)

Yoshida, Hiroyuki; Wu, Yin; Cai, Wenli

2014-03-01

One of the key challenges in three-dimensional (3D) medical imaging is to enable the fast turn-around time, which is often required for interactive or real-time response. This inevitably requires not only high computational power but also high memory bandwidth due to the massive amount of data that need to be processed. For this purpose, we previously developed a software platform for high-performance 3D medical image processing, called HPC 3D-MIP platform, which employs increasingly available and affordable commodity computing systems such as the multicore, cluster, and cloud computing systems. To achieve scalable high-performance computing, the platform employed size-adaptive, distributable block volumes as a core data structure for efficient parallelization of a wide range of 3D-MIP algorithms, supported task scheduling for efficient load distribution and balancing, and consisted of a layered parallel software libraries that allow image processing applications to share the common functionalities. We evaluated the performance of the HPC 3D-MIP platform by applying it to computationally intensive processes in virtual colonoscopy. Experimental results showed a 12-fold performance improvement on a workstation with 12-core CPUs over the original sequential implementation of the processes, indicating the efficiency of the platform. Analysis of performance scalability based on the Amdahl's law for symmetric multicore chips showed the potential of a high performance scalability of the HPC 3DMIP platform when a larger number of cores is available.

3D Kirchhoff depth migration algorithm: A new scalable approach for parallelization on multicore CPU based cluster

NASA Astrophysics Data System (ADS)

Rastogi, Richa; Londhe, Ashutosh; Srivastava, Abhishek; Sirasala, Kirannmayi M.; Khonde, Kiran

2017-03-01

In this article, a new scalable 3D Kirchhoff depth migration algorithm is presented on state of the art multicore CPU based cluster. Parallelization of 3D Kirchhoff depth migration is challenging due to its high demand of compute time, memory, storage and I/O along with the need of their effective management. The most resource intensive modules of the algorithm are traveltime calculations and migration summation which exhibit an inherent trade off between compute time and other resources. The parallelization strategy of the algorithm largely depends on the storage of calculated traveltimes and its feeding mechanism to the migration process. The presented work is an extension of our previous work, wherein a 3D Kirchhoff depth migration application for multicore CPU based parallel system had been developed. Recently, we have worked on improving parallel performance of this application by re-designing the parallelization approach. The new algorithm is capable to efficiently migrate both prestack and poststack 3D data. It exhibits flexibility for migrating large number of traces within the available node memory and with minimal requirement of storage, I/O and inter-node communication. The resultant application is tested using 3D Overthrust data on PARAM Yuva II, which is a Xeon E5-2670 based multicore CPU cluster with 16 cores/node and 64 GB shared memory. Parallel performance of the algorithm is studied using different numerical experiments and the scalability results show striking improvement over its previous version. An impressive 49.05X speedup with 76.64% efficiency is achieved for 3D prestack data and 32.00X speedup with 50.00% efficiency for 3D poststack data, using 64 nodes. The results also demonstrate the effectiveness and robustness of the improved algorithm with high scalability and efficiency on a multicore CPU cluster.

Development and Applications of a Modular Parallel Process for Large Scale Fluid/Structures Problems

NASA Technical Reports Server (NTRS)

Guruswamy, Guru P.; Kwak, Dochan (Technical Monitor)

2002-01-01

A modular process that can efficiently solve large scale multidisciplinary problems using massively parallel supercomputers is presented. The process integrates disciplines with diverse physical characteristics by retaining the efficiency of individual disciplines. Computational domain independence of individual disciplines is maintained using a meta programming approach. The process integrates disciplines without affecting the combined performance. Results are demonstrated for large scale aerospace problems on several supercomputers. The super scalability and portability of the approach is demonstrated on several parallel computers.

PIXIE3D: A Parallel, Implicit, eXtended MHD 3D Code

NASA Astrophysics Data System (ADS)

Chacon, Luis

2006-10-01

We report on the development of PIXIE3D, a 3D parallel, fully implicit Newton-Krylov extended MHD code in general curvilinear geometry. PIXIE3D employs a second-order, finite-volume-based spatial discretization that satisfies remarkable properties such as being conservative, solenoidal in the magnetic field to machine precision, non-dissipative, and linearly and nonlinearly stable in the absence of physical dissipation. PIXIE3D employs fully-implicit Newton-Krylov methods for the time advance. Currently, second-order implicit schemes such as Crank-Nicolson and BDF2 (2^nd order backward differentiation formula) are available. PIXIE3D is fully parallel (employs PETSc for parallelism), and exhibits excellent parallel scalability. A parallel, scalable, MG preconditioning strategy, based on physics-based preconditioning ideas, has been developed for resistive MHD, and is currently being extended to Hall MHD. In this poster, we will report on progress in the algorithmic formulation for extended MHD, as well as the the serial and parallel performance of PIXIE3D in a variety of problems and geometries. L. Chac'on, Comput. Phys. Comm., 163 (3), 143-171 (2004) L. Chac'on et al., J. Comput. Phys. 178 (1), 15- 36 (2002); J. Comput. Phys., 188 (2), 573-592 (2003) L. Chac'on, 32nd EPS Conf. Plasma Physics, Tarragona, Spain, 2005 L. Chac'on et al., 33rd EPS Conf. Plasma Physics, Rome, Italy, 2006

A novel processing platform for post tape out flows

NASA Astrophysics Data System (ADS)

Vu, Hien T.; Kim, Soohong; Word, James; Cai, Lynn Y.

2018-03-01

As the computational requirements for post tape out (PTO) flows increase at the 7nm and below technology nodes, there is a need to increase the scalability of the computational tools in order to reduce the turn-around time (TAT) of the flows. Utilization of design hierarchy has been one proven method to provide sufficient partitioning to enable PTO processing. However, as the data is processed through the PTO flow, its effective hierarchy is reduced. The reduction is necessary to achieve the desired accuracy. Also, the sequential nature of the PTO flow is inherently non-scalable. To address these limitations, we are proposing a quasi-hierarchical solution that combines multiple levels of parallelism to increase the scalability of the entire PTO flow. In this paper, we describe the system and present experimental results demonstrating the runtime reduction through scalable processing with thousands of computational cores.

: A Scalable and Transparent System for Simulating MPI Programs

DOE Office of Scientific and Technical Information (OSTI.GOV)

Perumalla, Kalyan S

2010-01-01

is a scalable, transparent system for experimenting with the execution of parallel programs on simulated computing platforms. The level of simulated detail can be varied for application behavior as well as for machine characteristics. Unique features of are repeatability of execution, scalability to millions of simulated (virtual) MPI ranks, scalability to hundreds of thousands of host (real) MPI ranks, portability of the system to a variety of host supercomputing platforms, and the ability to experiment with scientific applications whose source-code is available. The set of source-code interfaces supported by is being expanded to support a wider set of applications, andmore » MPI-based scientific computing benchmarks are being ported. In proof-of-concept experiments, has been successfully exercised to spawn and sustain very large-scale executions of an MPI test program given in source code form. Low slowdowns are observed, due to its use of purely discrete event style of execution, and due to the scalability and efficiency of the underlying parallel discrete event simulation engine, sik. In the largest runs, has been executed on up to 216,000 cores of a Cray XT5 supercomputer, successfully simulating over 27 million virtual MPI ranks, each virtual rank containing its own thread context, and all ranks fully synchronized by virtual time.« less

Scalable Parallel Computation for Extended MHD Modeling of Fusion Plasmas

NASA Astrophysics Data System (ADS)

Glasser, Alan H.

2008-11-01

Parallel solution of a linear system is scalable if simultaneously doubling the number of dependent variables and the number of processors results in little or no increase in the computation time to solution. Two approaches have this property for parabolic systems: multigrid and domain decomposition. Since extended MHD is primarily a hyperbolic rather than a parabolic system, additional steps must be taken to parabolize the linear system to be solved by such a method. Such physics-based preconditioning (PBP) methods have been pioneered by Chac'on, using finite volumes for spatial discretization, multigrid for solution of the preconditioning equations, and matrix-free Newton-Krylov methods for the accurate solution of the full nonlinear preconditioned equations. The work described here is an extension of these methods using high-order spectral element methods and FETI-DP domain decomposition. Application of PBP to a flux-source representation of the physics equations is discussed. The resulting scalability will be demonstrated for simple wave and for ideal and Hall MHD waves.

Implementation of the NAS Parallel Benchmarks in Java

NASA Technical Reports Server (NTRS)

Frumkin, Michael A.; Schultz, Matthew; Jin, Haoqiang; Yan, Jerry; Biegel, Bryan (Technical Monitor)

2002-01-01

Several features make Java an attractive choice for High Performance Computing (HPC). In order to gauge the applicability of Java to Computational Fluid Dynamics (CFD), we have implemented the NAS (NASA Advanced Supercomputing) Parallel Benchmarks in Java. The performance and scalability of the benchmarks point out the areas where improvement in Java compiler technology and in Java thread implementation would position Java closer to Fortran in the competition for CFD applications.

PIXIE3D: A Parallel, Implicit, eXtended MHD 3D Code.

NASA Astrophysics Data System (ADS)

Chacon, L.; Knoll, D. A.

2004-11-01

We report on the development of PIXIE3D, a 3D parallel, fully implicit Newton-Krylov extended primitive-variable MHD code in general curvilinear geometry. PIXIE3D employs a second-order, finite-volume-based spatial discretization that satisfies remarkable properties such as being conservative, solenoidal in the magnetic field, non-dissipative, and stable in the absence of physical dissipation.(L. Chacón , phComput. Phys. Comm.) submitted (2004) PIXIE3D employs fully-implicit Newton-Krylov methods for the time advance. Currently, first and second-order implicit schemes are available, although higher-order temporal implicit schemes can be effortlessly implemented within the Newton-Krylov framework. A successful, scalable, MG physics-based preconditioning strategy, similar in concept to previous 2D MHD efforts,(L. Chacón et al., phJ. Comput. Phys). 178 (1), 15- 36 (2002); phJ. Comput. Phys., 188 (2), 573-592 (2003) has been developed. We are currently in the process of parallelizing the code using the PETSc library, and a Newton-Krylov-Schwarz approach for the parallel treatment of the preconditioner. In this poster, we will report on both the serial and parallel performance of PIXIE3D, focusing primarily on scalability and CPU speedup vs. an explicit approach.

Scalability study of parallel spatial direct numerical simulation code on IBM SP1 parallel supercomputer

NASA Technical Reports Server (NTRS)

Hanebutte, Ulf R.; Joslin, Ronald D.; Zubair, Mohammad

1994-01-01

The implementation and the performance of a parallel spatial direct numerical simulation (PSDNS) code are reported for the IBM SP1 supercomputer. The spatially evolving disturbances that are associated with laminar-to-turbulent in three-dimensional boundary-layer flows are computed with the PS-DNS code. By remapping the distributed data structure during the course of the calculation, optimized serial library routines can be utilized that substantially increase the computational performance. Although the remapping incurs a high communication penalty, the parallel efficiency of the code remains above 40% for all performed calculations. By using appropriate compile options and optimized library routines, the serial code achieves 52-56 Mflops on a single node of the SP1 (45% of theoretical peak performance). The actual performance of the PSDNS code on the SP1 is evaluated with a 'real world' simulation that consists of 1.7 million grid points. One time step of this simulation is calculated on eight nodes of the SP1 in the same time as required by a Cray Y/MP for the same simulation. The scalability information provides estimated computational costs that match the actual costs relative to changes in the number of grid points.

Scalable splitting algorithms for big-data interferometric imaging in the SKA era

NASA Astrophysics Data System (ADS)

Onose, Alexandru; Carrillo, Rafael E.; Repetti, Audrey; McEwen, Jason D.; Thiran, Jean-Philippe; Pesquet, Jean-Christophe; Wiaux, Yves

2016-11-01

In the context of next-generation radio telescopes, like the Square Kilometre Array (SKA), the efficient processing of large-scale data sets is extremely important. Convex optimization tasks under the compressive sensing framework have recently emerged and provide both enhanced image reconstruction quality and scalability to increasingly larger data sets. We focus herein mainly on scalability and propose two new convex optimization algorithmic structures able to solve the convex optimization tasks arising in radio-interferometric imaging. They rely on proximal splitting and forward-backward iterations and can be seen, by analogy, with the CLEAN major-minor cycle, as running sophisticated CLEAN-like iterations in parallel in multiple data, prior, and image spaces. Both methods support any convex regularization function, in particular, the well-studied ℓ1 priors promoting image sparsity in an adequate domain. Tailored for big-data, they employ parallel and distributed computations to achieve scalability, in terms of memory and computational requirements. One of them also exploits randomization, over data blocks at each iteration, offering further flexibility. We present simulation results showing the feasibility of the proposed methods as well as their advantages compared to state-of-the-art algorithmic solvers. Our MATLAB code is available online on GitHub.

An O(N) and parallel approach to integral problems by a kernel-independent fast multipole method: Application to polarization and magnetization of interacting particles

NASA Astrophysics Data System (ADS)

Jiang, Xikai; Li, Jiyuan; Zhao, Xujun; Qin, Jian; Karpeev, Dmitry; Hernandez-Ortiz, Juan; de Pablo, Juan J.; Heinonen, Olle

2016-08-01

Large classes of materials systems in physics and engineering are governed by magnetic and electrostatic interactions. Continuum or mesoscale descriptions of such systems can be cast in terms of integral equations, whose direct computational evaluation requires O(N2) operations, where N is the number of unknowns. Such a scaling, which arises from the many-body nature of the relevant Green's function, has precluded wide-spread adoption of integral methods for solution of large-scale scientific and engineering problems. In this work, a parallel computational approach is presented that relies on using scalable open source libraries and utilizes a kernel-independent Fast Multipole Method (FMM) to evaluate the integrals in O(N) operations, with O(N) memory cost, thereby substantially improving the scalability and efficiency of computational integral methods. We demonstrate the accuracy, efficiency, and scalability of our approach in the context of two examples. In the first, we solve a boundary value problem for a ferroelectric/ferromagnetic volume in free space. In the second, we solve an electrostatic problem involving polarizable dielectric bodies in an unbounded dielectric medium. The results from these test cases show that our proposed parallel approach, which is built on a kernel-independent FMM, can enable highly efficient and accurate simulations and allow for considerable flexibility in a broad range of applications.

An O( N) and parallel approach to integral problems by a kernel-independent fast multipole method: Application to polarization and magnetization of interacting particles

DOE PAGES

Jiang, Xikai; Li, Jiyuan; Zhao, Xujun; ...

2016-08-10

Large classes of materials systems in physics and engineering are governed by magnetic and electrostatic interactions. Continuum or mesoscale descriptions of such systems can be cast in terms of integral equations, whose direct computational evaluation requires O( N 2) operations, where N is the number of unknowns. Such a scaling, which arises from the many-body nature of the relevant Green's function, has precluded wide-spread adoption of integral methods for solution of large-scale scientific and engineering problems. In this work, a parallel computational approach is presented that relies on using scalable open source libraries and utilizes a kernel-independent Fast Multipole Methodmore » (FMM) to evaluate the integrals in O( N) operations, with O( N) memory cost, thereby substantially improving the scalability and efficiency of computational integral methods. We demonstrate the accuracy, efficiency, and scalability of our approach in the context of two examples. In the first, we solve a boundary value problem for a ferroelectric/ferromagnetic volume in free space. In the second, we solve an electrostatic problem involving polarizable dielectric bodies in an unbounded dielectric medium. Lastly, the results from these test cases show that our proposed parallel approach, which is built on a kernel-independent FMM, can enable highly efficient and accurate simulations and allow for considerable flexibility in a broad range of applications.« less

GASPRNG: GPU accelerated scalable parallel random number generator library

NASA Astrophysics Data System (ADS)

Gao, Shuang; Peterson, Gregory D.

2013-04-01

Graphics processors represent a promising technology for accelerating computational science applications. Many computational science applications require fast and scalable random number generation with good statistical properties, so they use the Scalable Parallel Random Number Generators library (SPRNG). We present the GPU Accelerated SPRNG library (GASPRNG) to accelerate SPRNG in GPU-based high performance computing systems. GASPRNG includes code for a host CPU and CUDA code for execution on NVIDIA graphics processing units (GPUs) along with a programming interface to support various usage models for pseudorandom numbers and computational science applications executing on the CPU, GPU, or both. This paper describes the implementation approach used to produce high performance and also describes how to use the programming interface. The programming interface allows a user to be able to use GASPRNG the same way as SPRNG on traditional serial or parallel computers as well as to develop tightly coupled programs executing primarily on the GPU. We also describe how to install GASPRNG and use it. To help illustrate linking with GASPRNG, various demonstration codes are included for the different usage models. GASPRNG on a single GPU shows up to 280x speedup over SPRNG on a single CPU core and is able to scale for larger systems in the same manner as SPRNG. Because GASPRNG generates identical streams of pseudorandom numbers as SPRNG, users can be confident about the quality of GASPRNG for scalable computational science applications. Catalogue identifier: AEOI_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEOI_v1_0.html Program obtainable from: CPC Program Library, Queen’s University, Belfast, N. Ireland Licensing provisions: UTK license. No. of lines in distributed program, including test data, etc.: 167900 No. of bytes in distributed program, including test data, etc.: 1422058 Distribution format: tar.gz Programming language: C and CUDA. Computer: Any PC or workstation with NVIDIA GPU (Tested on Fermi GTX480, Tesla C1060, Tesla M2070). Operating system: Linux with CUDA version 4.0 or later. Should also run on MacOS, Windows, or UNIX. Has the code been vectorized or parallelized?: Yes. Parallelized using MPI directives. RAM: 512 MB˜ 732 MB (main memory on host CPU, depending on the data type of random numbers.) / 512 MB (GPU global memory) Classification: 4.13, 6.5. Nature of problem: Many computational science applications are able to consume large numbers of random numbers. For example, Monte Carlo simulations are able to consume limitless random numbers for the computation as long as resources for the computing are supported. Moreover, parallel computational science applications require independent streams of random numbers to attain statistically significant results. The SPRNG library provides this capability, but at a significant computational cost. The GASPRNG library presented here accelerates the generators of independent streams of random numbers using graphical processing units (GPUs). Solution method: Multiple copies of random number generators in GPUs allow a computational science application to consume large numbers of random numbers from independent, parallel streams. GASPRNG is a random number generators library to allow a computational science application to employ multiple copies of random number generators to boost performance. Users can interface GASPRNG with software code executing on microprocessors and/or GPUs. Running time: The tests provided take a few minutes to run.

A scalable PC-based parallel computer for lattice QCD

NASA Astrophysics Data System (ADS)

Fodor, Z.; Katz, S. D.; Pappa, G.

2003-05-01

A PC-based parallel computer for medium/large scale lattice QCD simulations is suggested. The Eo¨tvo¨s Univ., Inst. Theor. Phys. cluster consists of 137 Intel P4-1.7GHz nodes. Gigabit Ethernet cards are used for nearest neighbor communication in a two-dimensional mesh. The sustained performance for dynamical staggered (wilson) quarks on large lattices is around 70(110) GFlops. The exceptional price/performance ratio is below $1/Mflop.

Implementation of NAS Parallel Benchmarks in Java

NASA Technical Reports Server (NTRS)

Frumkin, Michael; Schultz, Matthew; Jin, Hao-Qiang; Yan, Jerry

2000-01-01

A number of features make Java an attractive but a debatable choice for High Performance Computing (HPC). In order to gauge the applicability of Java to the Computational Fluid Dynamics (CFD) we have implemented NAS Parallel Benchmarks in Java. The performance and scalability of the benchmarks point out the areas where improvement in Java compiler technology and in Java thread implementation would move Java closer to Fortran in the competition for CFD applications.

Scalable Parallel Algorithms for Multidimensional Digital Signal Processing

DTIC Science & Technology

1991-12-31

Proceedings, San Diego CL., August 1989, pp. 132-146. 53 [13] A. L. Gorin, L. Auslander, and A. Silberger . Balanced computation of 2D trans- forms on a tree...Speech, Signal Processing. ASSP-34, Oct. 1986,pp. 1301-1309. [24] A. Norton and A. Silberger . Parallelization and performance analysis of the Cooley-Tukey

Profiling and Improving I/O Performance of a Large-Scale Climate Scientific Application

NASA Technical Reports Server (NTRS)

Liu, Zhuo; Wang, Bin; Wang, Teng; Tian, Yuan; Xu, Cong; Wang, Yandong; Yu, Weikuan; Cruz, Carlos A.; Zhou, Shujia; Clune, Tom;

Highly Scalable Asynchronous Computing Method for Partial Differential Equations: A Path Towards Exascale

NASA Astrophysics Data System (ADS)

Konduri, Aditya

Many natural and engineering systems are governed by nonlinear partial differential equations (PDEs) which result in a multiscale phenomena, e.g. turbulent flows. Numerical simulations of these problems are computationally very expensive and demand for extreme levels of parallelism. At realistic conditions, simulations are being carried out on massively parallel computers with hundreds of thousands of processing elements (PEs). It has been observed that communication between PEs as well as their synchronization at these extreme scales take up a significant portion of the total simulation time and result in poor scalability of codes. This issue is likely to pose a bottleneck in scalability of codes on future Exascale systems. In this work, we propose an asynchronous computing algorithm based on widely used finite difference methods to solve PDEs in which synchronization between PEs due to communication is relaxed at a mathematical level. We show that while stability is conserved when schemes are used asynchronously, accuracy is greatly degraded. Since message arrivals at PEs are random processes, so is the behavior of the error. We propose a new statistical framework in which we show that average errors drop always to first-order regardless of the original scheme. We propose new asynchrony-tolerant schemes that maintain accuracy when synchronization is relaxed. The quality of the solution is shown to depend, not only on the physical phenomena and numerical schemes, but also on the characteristics of the computing machine. A novel algorithm using remote memory access communications has been developed to demonstrate excellent scalability of the method for large-scale computing. Finally, we present a path to extend this method in solving complex multi-scale problems on Exascale machines.

A Framework for Load Balancing of Tensor Contraction Expressions via Dynamic Task Partitioning

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lai, Pai-Wei; Stock, Kevin; Rajbhandari, Samyam

In this paper, we introduce the Dynamic Load-balanced Tensor Contractions (DLTC), a domain-specific library for efficient task parallel execution of tensor contraction expressions, a class of computation encountered in quantum chemistry and physics. Our framework decomposes each contraction into smaller unit of tasks, represented by an abstraction referred to as iterators. We exploit an extra level of parallelism by having tasks across independent contractions executed concurrently through a dynamic load balancing run- time. We demonstrate the improved performance, scalability, and flexibility for the computation of tensor contraction expressions on parallel computers using examples from coupled cluster methods.

Parallel and Scalable Clustering and Classification for Big Data in Geosciences

NASA Astrophysics Data System (ADS)

Riedel, M.

2015-12-01

Machine learning, data mining, and statistical computing are common techniques to perform analysis in earth sciences. This contribution will focus on two concrete and widely used data analytics methods suitable to analyse 'big data' in the context of geoscience use cases: clustering and classification. From the broad class of available clustering methods we focus on the density-based spatial clustering of appliactions with noise (DBSCAN) algorithm that enables the identification of outliers or interesting anomalies. A new open source parallel and scalable DBSCAN implementation will be discussed in the light of a scientific use case that detects water mixing events in the Koljoefjords. The second technique we cover is classification, with a focus set on the support vector machines algorithm (SVMs), as one of the best out-of-the-box classification algorithm. A parallel and scalable SVM implementation will be discussed in the light of a scientific use case in the field of remote sensing with 52 different classes of land cover types.

High performance data transfer

NASA Astrophysics Data System (ADS)

Cottrell, R.; Fang, C.; Hanushevsky, A.; Kreuger, W.; Yang, W.

2017-10-01

The exponentially increasing need for high speed data transfer is driven by big data, and cloud computing together with the needs of data intensive science, High Performance Computing (HPC), defense, the oil and gas industry etc. We report on the Zettar ZX software. This has been developed since 2013 to meet these growing needs by providing high performance data transfer and encryption in a scalable, balanced, easy to deploy and use way while minimizing power and space utilization. In collaboration with several commercial vendors, Proofs of Concept (PoC) consisting of clusters have been put together using off-the- shelf components to test the ZX scalability and ability to balance services using multiple cores, and links. The PoCs are based on SSD flash storage that is managed by a parallel file system. Each cluster occupies 4 rack units. Using the PoCs, between clusters we have achieved almost 200Gbps memory to memory over two 100Gbps links, and 70Gbps parallel file to parallel file with encryption over a 5000 mile 100Gbps link.

Automated Performance Prediction of Message-Passing Parallel Programs

NASA Technical Reports Server (NTRS)

Block, Robert J.; Sarukkai, Sekhar; Mehra, Pankaj; Woodrow, Thomas S. (Technical Monitor)

1995-01-01

The increasing use of massively parallel supercomputers to solve large-scale scientific problems has generated a need for tools that can predict scalability trends of applications written for these machines. Much work has been done to create simple models that represent important characteristics of parallel programs, such as latency, network contention, and communication volume. But many of these methods still require substantial manual effort to represent an application in the model's format. The NIK toolkit described in this paper is the result of an on-going effort to automate the formation of analytic expressions of program execution time, with a minimum of programmer assistance. In this paper we demonstrate the feasibility of our approach, by extending previous work to detect and model communication patterns automatically, with and without overlapped computations. The predictions derived from these models agree, within reasonable limits, with execution times of programs measured on the Intel iPSC/860 and Paragon. Further, we demonstrate the use of MK in selecting optimal computational grain size and studying various scalability metrics.

Parallel rendering

NASA Technical Reports Server (NTRS)

Crockett, Thomas W.

1995-01-01

This article provides a broad introduction to the subject of parallel rendering, encompassing both hardware and software systems. The focus is on the underlying concepts and the issues which arise in the design of parallel rendering algorithms and systems. We examine the different types of parallelism and how they can be applied in rendering applications. Concepts from parallel computing, such as data decomposition, task granularity, scalability, and load balancing, are considered in relation to the rendering problem. We also explore concepts from computer graphics, such as coherence and projection, which have a significant impact on the structure of parallel rendering algorithms. Our survey covers a number of practical considerations as well, including the choice of architectural platform, communication and memory requirements, and the problem of image assembly and display. We illustrate the discussion with numerous examples from the parallel rendering literature, representing most of the principal rendering methods currently used in computer graphics.

Scalable and balanced dynamic hybrid data assimilation

NASA Astrophysics Data System (ADS)

Kauranne, Tuomo; Amour, Idrissa; Gunia, Martin; Kallio, Kari; Lepistö, Ahti; Koponen, Sampsa

2017-04-01

Scalability of complex weather forecasting suites is dependent on the technical tools available for implementing highly parallel computational kernels, but to an equally large extent also on the dependence patterns between various components of the suite, such as observation processing, data assimilation and the forecast model. Scalability is a particular challenge for 4D variational assimilation methods that necessarily couple the forecast model into the assimilation process and subject this combination to an inherently serial quasi-Newton minimization process. Ensemble based assimilation methods are naturally more parallel, but large models force ensemble sizes to be small and that results in poor assimilation accuracy, somewhat akin to shooting with a shotgun in a million-dimensional space. The Variational Ensemble Kalman Filter (VEnKF) is an ensemble method that can attain the accuracy of 4D variational data assimilation with a small ensemble size. It achieves this by processing a Gaussian approximation of the current error covariance distribution, instead of a set of ensemble members, analogously to the Extended Kalman Filter EKF. Ensemble members are re-sampled every time a new set of observations is processed from a new approximation of that Gaussian distribution which makes VEnKF a dynamic assimilation method. After this a smoothing step is applied that turns VEnKF into a dynamic Variational Ensemble Kalman Smoother VEnKS. In this smoothing step, the same process is iterated with frequent re-sampling of the ensemble but now using past iterations as surrogate observations until the end result is a smooth and balanced model trajectory. In principle, VEnKF could suffer from similar scalability issues as 4D-Var. However, this can be avoided by isolating the forecast model completely from the minimization process by implementing the latter as a wrapper code whose only link to the model is calling for many parallel and totally independent model runs, all of them implemented as parallel model runs themselves. The only bottleneck in the process is the gathering and scattering of initial and final model state snapshots before and after the parallel runs which requires a very efficient and low-latency communication network. However, the volume of data communicated is small and the intervening minimization steps are only 3D-Var, which means their computational load is negligible compared with the fully parallel model runs. We present example results of scalable VEnKF with the 4D lake and shallow sea model COHERENS, assimilating simultaneously continuous in situ measurements in a single point and infrequent satellite images that cover a whole lake, with the fully scalable VEnKF.

Ultrafast and scalable cone-beam CT reconstruction using MapReduce in a cloud computing environment.

PubMed

Meng, Bowen; Pratx, Guillem; Xing, Lei

2011-12-01

Four-dimensional CT (4DCT) and cone beam CT (CBCT) are widely used in radiation therapy for accurate tumor target definition and localization. However, high-resolution and dynamic image reconstruction is computationally demanding because of the large amount of data processed. Efficient use of these imaging techniques in the clinic requires high-performance computing. The purpose of this work is to develop a novel ultrafast, scalable and reliable image reconstruction technique for 4D CBCT∕CT using a parallel computing framework called MapReduce. We show the utility of MapReduce for solving large-scale medical physics problems in a cloud computing environment. In this work, we accelerated the Feldcamp-Davis-Kress (FDK) algorithm by porting it to Hadoop, an open-source MapReduce implementation. Gated phases from a 4DCT scans were reconstructed independently. Following the MapReduce formalism, Map functions were used to filter and backproject subsets of projections, and Reduce function to aggregate those partial backprojection into the whole volume. MapReduce automatically parallelized the reconstruction process on a large cluster of computer nodes. As a validation, reconstruction of a digital phantom and an acquired CatPhan 600 phantom was performed on a commercial cloud computing environment using the proposed 4D CBCT∕CT reconstruction algorithm. Speedup of reconstruction time is found to be roughly linear with the number of nodes employed. For instance, greater than 10 times speedup was achieved using 200 nodes for all cases, compared to the same code executed on a single machine. Without modifying the code, faster reconstruction is readily achievable by allocating more nodes in the cloud computing environment. Root mean square error between the images obtained using MapReduce and a single-threaded reference implementation was on the order of 10(-7). Our study also proved that cloud computing with MapReduce is fault tolerant: the reconstruction completed successfully with identical results even when half of the nodes were manually terminated in the middle of the process. An ultrafast, reliable and scalable 4D CBCT∕CT reconstruction method was developed using the MapReduce framework. Unlike other parallel computing approaches, the parallelization and speedup required little modification of the original reconstruction code. MapReduce provides an efficient and fault tolerant means of solving large-scale computing problems in a cloud computing environment.

Ultrafast and scalable cone-beam CT reconstruction using MapReduce in a cloud computing environment

PubMed Central

Meng, Bowen; Pratx, Guillem; Xing, Lei

2011-01-01

Purpose: Four-dimensional CT (4DCT) and cone beam CT (CBCT) are widely used in radiation therapy for accurate tumor target definition and localization. However, high-resolution and dynamic image reconstruction is computationally demanding because of the large amount of data processed. Efficient use of these imaging techniques in the clinic requires high-performance computing. The purpose of this work is to develop a novel ultrafast, scalable and reliable image reconstruction technique for 4D CBCT/CT using a parallel computing framework called MapReduce. We show the utility of MapReduce for solving large-scale medical physics problems in a cloud computing environment. Methods: In this work, we accelerated the Feldcamp–Davis–Kress (FDK) algorithm by porting it to Hadoop, an open-source MapReduce implementation. Gated phases from a 4DCT scans were reconstructed independently. Following the MapReduce formalism, Map functions were used to filter and backproject subsets of projections, and Reduce function to aggregate those partial backprojection into the whole volume. MapReduce automatically parallelized the reconstruction process on a large cluster of computer nodes. As a validation, reconstruction of a digital phantom and an acquired CatPhan 600 phantom was performed on a commercial cloud computing environment using the proposed 4D CBCT/CT reconstruction algorithm. Results: Speedup of reconstruction time is found to be roughly linear with the number of nodes employed. For instance, greater than 10 times speedup was achieved using 200 nodes for all cases, compared to the same code executed on a single machine. Without modifying the code, faster reconstruction is readily achievable by allocating more nodes in the cloud computing environment. Root mean square error between the images obtained using MapReduce and a single-threaded reference implementation was on the order of 10−7. Our study also proved that cloud computing with MapReduce is fault tolerant: the reconstruction completed successfully with identical results even when half of the nodes were manually terminated in the middle of the process. Conclusions: An ultrafast, reliable and scalable 4D CBCT/CT reconstruction method was developed using the MapReduce framework. Unlike other parallel computing approaches, the parallelization and speedup required little modification of the original reconstruction code. MapReduce provides an efficient and fault tolerant means of solving large-scale computing problems in a cloud computing environment. PMID:22149842

MPI implementation of PHOENICS: A general purpose computational fluid dynamics code

NASA Astrophysics Data System (ADS)

Simunovic, S.; Zacharia, T.; Baltas, N.; Spalding, D. B.

1995-03-01

PHOENICS is a suite of computational analysis programs that are used for simulation of fluid flow, heat transfer, and dynamical reaction processes. The parallel version of the solver EARTH for the Computational Fluid Dynamics (CFD) program PHOENICS has been implemented using Message Passing Interface (MPI) standard. Implementation of MPI version of PHOENICS makes this computational tool portable to a wide range of parallel machines and enables the use of high performance computing for large scale computational simulations. MPI libraries are available on several parallel architectures making the program usable across different architectures as well as on heterogeneous computer networks. The Intel Paragon NX and MPI versions of the program have been developed and tested on massively parallel supercomputers Intel Paragon XP/S 5, XP/S 35, and Kendall Square Research, and on the multiprocessor SGI Onyx computer at Oak Ridge National Laboratory. The preliminary testing results of the developed program have shown scalable performance for reasonably sized computational domains.

MPI implementation of PHOENICS: A general purpose computational fluid dynamics code

DOE Office of Scientific and Technical Information (OSTI.GOV)

Simunovic, S.; Zacharia, T.; Baltas, N.

1995-04-01

PHOENICS is a suite of computational analysis programs that are used for simulation of fluid flow, heat transfer, and dynamical reaction processes. The parallel version of the solver EARTH for the Computational Fluid Dynamics (CFD) program PHOENICS has been implemented using Message Passing Interface (MPI) standard. Implementation of MPI version of PHOENICS makes this computational tool portable to a wide range of parallel machines and enables the use of high performance computing for large scale computational simulations. MPI libraries are available on several parallel architectures making the program usable across different architectures as well as on heterogeneous computer networks. Themore » Intel Paragon NX and MPI versions of the program have been developed and tested on massively parallel supercomputers Intel Paragon XP/S 5, XP/S 35, and Kendall Square Research, and on the multiprocessor SGI Onyx computer at Oak Ridge National Laboratory. The preliminary testing results of the developed program have shown scalable performance for reasonably sized computational domains.« less

Argonne Simulation Framework for Intelligent Transportation Systems

DOT National Transportation Integrated Search

1996-01-01

A simulation framework has been developed which defines a high-level architecture for a large-scale, comprehensive, scalable simulation of an Intelligent Transportation System (ITS). The simulator is designed to run on parallel computers and distribu...

Scalable Computing of the Mesh Size Effect on Modeling Damage Mechanics in Woven Armor Composites

DTIC Science & Technology

2008-12-01

manner of a user defined material subroutine to provide overall stress increments to, the parallel LS-DYNA3D a Lagrangian explicit code used in...finite element code, as a user defined material subroutine . The ability of this subroutine to model the effect of the progressions of a select number...is added as a user defined material subroutine to parallel LS-DYNA3D. The computations of the global mesh are handled by LS-DYNA3D and are spread

Field-Programmable Gate Array Computer in Structural Analysis: An Initial Exploration

NASA Technical Reports Server (NTRS)

Singleterry, Robert C., Jr.; Sobieszczanski-Sobieski, Jaroslaw; Brown, Samuel

2002-01-01

This paper reports on an initial assessment of using a Field-Programmable Gate Array (FPGA) computational device as a new tool for solving structural mechanics problems. A FPGA is an assemblage of binary gates arranged in logical blocks that are interconnected via software in a manner dependent on the algorithm being implemented and can be reprogrammed thousands of times per second. In effect, this creates a computer specialized for the problem that automatically exploits all the potential for parallel computing intrinsic in an algorithm. This inherent parallelism is the most important feature of the FPGA computational environment. It is therefore important that if a problem offers a choice of different solution algorithms, an algorithm of a higher degree of inherent parallelism should be selected. It is found that in structural analysis, an 'analog computer' style of programming, which solves problems by direct simulation of the terms in the governing differential equations, yields a more favorable solution algorithm than current solution methods. This style of programming is facilitated by a 'drag-and-drop' graphic programming language that is supplied with the particular type of FPGA computer reported in this paper. Simple examples in structural dynamics and statics illustrate the solution approach used. The FPGA system also allows linear scalability in computing capability. As the problem grows, the number of FPGA chips can be increased with no loss of computing efficiency due to data flow or algorithmic latency that occurs when a single problem is distributed among many conventional processors that operate in parallel. This initial assessment finds the FPGA hardware and software to be in their infancy in regard to the user conveniences; however, they have enormous potential for shrinking the elapsed time of structural analysis solutions if programmed with algorithms that exhibit inherent parallelism and linear scalability. This potential warrants further development of FPGA-tailored algorithms for structural analysis.

Large-scale three-dimensional phase-field simulations for phase coarsening at ultrahigh volume fraction on high-performance architectures

NASA Astrophysics Data System (ADS)

Yan, Hui; Wang, K. G.; Jones, Jim E.

2016-06-01

A parallel algorithm for large-scale three-dimensional phase-field simulations of phase coarsening is developed and implemented on high-performance architectures. From the large-scale simulations, a new kinetics in phase coarsening in the region of ultrahigh volume fraction is found. The parallel implementation is capable of harnessing the greater computer power available from high-performance architectures. The parallelized code enables increase in three-dimensional simulation system size up to a 5123 grid cube. Through the parallelized code, practical runtime can be achieved for three-dimensional large-scale simulations, and the statistical significance of the results from these high resolution parallel simulations are greatly improved over those obtainable from serial simulations. A detailed performance analysis on speed-up and scalability is presented, showing good scalability which improves with increasing problem size. In addition, a model for prediction of runtime is developed, which shows a good agreement with actual run time from numerical tests.

The ELPA library: scalable parallel eigenvalue solutions for electronic structure theory and computational science.

PubMed

Marek, A; Blum, V; Johanni, R; Havu, V; Lang, B; Auckenthaler, T; Heinecke, A; Bungartz, H-J; Lederer, H

2014-05-28

Obtaining the eigenvalues and eigenvectors of large matrices is a key problem in electronic structure theory and many other areas of computational science. The computational effort formally scales as O(N(3)) with the size of the investigated problem, N (e.g. the electron count in electronic structure theory), and thus often defines the system size limit that practical calculations cannot overcome. In many cases, more than just a small fraction of the possible eigenvalue/eigenvector pairs is needed, so that iterative solution strategies that focus only on a few eigenvalues become ineffective. Likewise, it is not always desirable or practical to circumvent the eigenvalue solution entirely. We here review some current developments regarding dense eigenvalue solvers and then focus on the Eigenvalue soLvers for Petascale Applications (ELPA) library, which facilitates the efficient algebraic solution of symmetric and Hermitian eigenvalue problems for dense matrices that have real-valued and complex-valued matrix entries, respectively, on parallel computer platforms. ELPA addresses standard as well as generalized eigenvalue problems, relying on the well documented matrix layout of the Scalable Linear Algebra PACKage (ScaLAPACK) library but replacing all actual parallel solution steps with subroutines of its own. For these steps, ELPA significantly outperforms the corresponding ScaLAPACK routines and proprietary libraries that implement the ScaLAPACK interface (e.g. Intel's MKL). The most time-critical step is the reduction of the matrix to tridiagonal form and the corresponding backtransformation of the eigenvectors. ELPA offers both a one-step tridiagonalization (successive Householder transformations) and a two-step transformation that is more efficient especially towards larger matrices and larger numbers of CPU cores. ELPA is based on the MPI standard, with an early hybrid MPI-OpenMPI implementation available as well. Scalability beyond 10,000 CPU cores for problem sizes arising in the field of electronic structure theory is demonstrated for current high-performance computer architectures such as Cray or Intel/Infiniband. For a matrix of dimension 260,000, scalability up to 295,000 CPU cores has been shown on BlueGene/P.

Linear static structural and vibration analysis on high-performance computers

NASA Technical Reports Server (NTRS)

Baddourah, M. A.; Storaasli, O. O.; Bostic, S. W.

1993-01-01

Parallel computers offer the oppurtunity to significantly reduce the computation time necessary to analyze large-scale aerospace structures. This paper presents algorithms developed for and implemented on massively-parallel computers hereafter referred to as Scalable High-Performance Computers (SHPC), for the most computationally intensive tasks involved in structural analysis, namely, generation and assembly of system matrices, solution of systems of equations and calculation of the eigenvalues and eigenvectors. Results on SHPC are presented for large-scale structural problems (i.e. models for High-Speed Civil Transport). The goal of this research is to develop a new, efficient technique which extends structural analysis to SHPC and makes large-scale structural analyses tractable.

A look at scalable dense linear algebra libraries

DOE Office of Scientific and Technical Information (OSTI.GOV)

Dongarra, J.J.; Van de Geijn, R.A.; Walker, D.W.

1992-01-01

We discuss the essential design features of a library of scalable software for performing dense linear algebra computations on distributed memory concurrent computers. The square block scattered decomposition is proposed as a flexible and general-purpose way of decomposing most, if not all, dense matrix problems. An object- oriented interface to the library permits more portable applications to be written, and is easy to learn and use, since details of the parallel implementation are hidden from the user. Experiments on the Intel Touchstone Delta system with a prototype code that uses the square block scattered decomposition to perform LU factorization aremore » presented and analyzed. It was found that the code was both scalable and efficient, performing at about 14 GFLOPS (double precision) for the largest problem considered.« less

A look at scalable dense linear algebra libraries

DOE Office of Scientific and Technical Information (OSTI.GOV)

Dongarra, J.J.; Van de Geijn, R.A.; Walker, D.W.

1992-08-01

We discuss the essential design features of a library of scalable software for performing dense linear algebra computations on distributed memory concurrent computers. The square block scattered decomposition is proposed as a flexible and general-purpose way of decomposing most, if not all, dense matrix problems. An object- oriented interface to the library permits more portable applications to be written, and is easy to learn and use, since details of the parallel implementation are hidden from the user. Experiments on the Intel Touchstone Delta system with a prototype code that uses the square block scattered decomposition to perform LU factorization aremore » presented and analyzed. It was found that the code was both scalable and efficient, performing at about 14 GFLOPS (double precision) for the largest problem considered.« less

A new augmentation based algorithm for extracting maximal chordal subgraphs

DOE PAGES

Bhowmick, Sanjukta; Chen, Tzu-Yi; Halappanavar, Mahantesh

2014-10-18

If every cycle of a graph is chordal length greater than three then it contains an edge between non-adjacent vertices. Chordal graphs are of interest both theoretically, since they admit polynomial time solutions to a range of NP-hard graph problems, and practically, since they arise in many applications including sparse linear algebra, computer vision, and computational biology. A maximal chordal subgraph is a chordal subgraph that is not a proper subgraph of any other chordal subgraph. Existing algorithms for computing maximal chordal subgraphs depend on dynamically ordering the vertices, which is an inherently sequential process and therefore limits the algorithms’more » parallelizability. In our paper we explore techniques to develop a scalable parallel algorithm for extracting a maximal chordal subgraph. We demonstrate that an earlier attempt at developing a parallel algorithm may induce a non-optimal vertex ordering and is therefore not guaranteed to terminate with a maximal chordal subgraph. We then give a new algorithm that first computes and then repeatedly augments a spanning chordal subgraph. After proving that the algorithm terminates with a maximal chordal subgraph, we then demonstrate that this algorithm is more amenable to parallelization and that the parallel version also terminates with a maximal chordal subgraph. That said, the complexity of the new algorithm is higher than that of the previous parallel algorithm, although the earlier algorithm computes a chordal subgraph which is not guaranteed to be maximal. Finally, we experimented with our augmentation-based algorithm on both synthetic and real-world graphs. We provide scalability results and also explore the effect of different choices for the initial spanning chordal subgraph on both the running time and on the number of edges in the maximal chordal subgraph.« less

A New Augmentation Based Algorithm for Extracting Maximal Chordal Subgraphs.

PubMed

Bhowmick, Sanjukta; Chen, Tzu-Yi; Halappanavar, Mahantesh

2015-02-01

A graph is chordal if every cycle of length greater than three contains an edge between non-adjacent vertices. Chordal graphs are of interest both theoretically, since they admit polynomial time solutions to a range of NP-hard graph problems, and practically, since they arise in many applications including sparse linear algebra, computer vision, and computational biology. A maximal chordal subgraph is a chordal subgraph that is not a proper subgraph of any other chordal subgraph. Existing algorithms for computing maximal chordal subgraphs depend on dynamically ordering the vertices, which is an inherently sequential process and therefore limits the algorithms' parallelizability. In this paper we explore techniques to develop a scalable parallel algorithm for extracting a maximal chordal subgraph. We demonstrate that an earlier attempt at developing a parallel algorithm may induce a non-optimal vertex ordering and is therefore not guaranteed to terminate with a maximal chordal subgraph. We then give a new algorithm that first computes and then repeatedly augments a spanning chordal subgraph. After proving that the algorithm terminates with a maximal chordal subgraph, we then demonstrate that this algorithm is more amenable to parallelization and that the parallel version also terminates with a maximal chordal subgraph. That said, the complexity of the new algorithm is higher than that of the previous parallel algorithm, although the earlier algorithm computes a chordal subgraph which is not guaranteed to be maximal. We experimented with our augmentation-based algorithm on both synthetic and real-world graphs. We provide scalability results and also explore the effect of different choices for the initial spanning chordal subgraph on both the running time and on the number of edges in the maximal chordal subgraph.

The fusion code XGC: Enabling kinetic study of multi-scale edge turbulent transport in ITER [Book Chapter

DOE Office of Scientific and Technical Information (OSTI.GOV)

D'Azevedo, Eduardo; Abbott, Stephen; Koskela, Tuomas

The XGC fusion gyrokinetic code combines state-of-the-art, portable computational and algorithmic technologies to enable complicated multiscale simulations of turbulence and transport dynamics in ITER edge plasma on the largest US open-science computer, the CRAY XK7 Titan, at its maximal heterogeneous capability, which have not been possible before due to a factor of over 10 shortage in the time-to-solution for less than 5 days of wall-clock time for one physics case. Frontier techniques such as nested OpenMP parallelism, adaptive parallel I/O, staging I/O and data reduction using dynamic and asynchronous applications interactions, dynamic repartitioning for balancing computational work in pushing particlesmore » and in grid related work, scalable and accurate discretization algorithms for non-linear Coulomb collisions, and communication-avoiding subcycling technology for pushing particles on both CPUs and GPUs are also utilized to dramatically improve the scalability and time-to-solution, hence enabling the difficult kinetic ITER edge simulation on a present-day leadership class computer.« less

Performance and scalability evaluation of "Big Memory" on Blue Gene Linux.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yoshii, K.; Iskra, K.; Naik, H.

2011-05-01

We address memory performance issues observed in Blue Gene Linux and discuss the design and implementation of 'Big Memory' - an alternative, transparent memory space introduced to eliminate the memory performance issues. We evaluate the performance of Big Memory using custom memory benchmarks, NAS Parallel Benchmarks, and the Parallel Ocean Program, at a scale of up to 4,096 nodes. We find that Big Memory successfully resolves the performance issues normally encountered in Blue Gene Linux. For the ocean simulation program, we even find that Linux with Big Memory provides better scalability than does the lightweight compute node kernel designed solelymore » for high-performance applications. Originally intended exclusively for compute node tasks, our new memory subsystem dramatically improves the performance of certain I/O node applications as well. We demonstrate this performance using the central processor of the LOw Frequency ARray radio telescope as an example.« less

Scalability of Parallel Spatial Direct Numerical Simulations on Intel Hypercube and IBM SP1 and SP2

NASA Technical Reports Server (NTRS)

Joslin, Ronald D.; Hanebutte, Ulf R.; Zubair, Mohammad

1995-01-01

The implementation and performance of a parallel spatial direct numerical simulation (PSDNS) approach on the Intel iPSC/860 hypercube and IBM SP1 and SP2 parallel computers is documented. Spatially evolving disturbances associated with the laminar-to-turbulent transition in boundary-layer flows are computed with the PSDNS code. The feasibility of using the PSDNS to perform transition studies on these computers is examined. The results indicate that PSDNS approach can effectively be parallelized on a distributed-memory parallel machine by remapping the distributed data structure during the course of the calculation. Scalability information is provided to estimate computational costs to match the actual costs relative to changes in the number of grid points. By increasing the number of processors, slower than linear speedups are achieved with optimized (machine-dependent library) routines. This slower than linear speedup results because the computational cost is dominated by FFT routine, which yields less than ideal speedups. By using appropriate compile options and optimized library routines on the SP1, the serial code achieves 52-56 M ops on a single node of the SP1 (45 percent of theoretical peak performance). The actual performance of the PSDNS code on the SP1 is evaluated with a "real world" simulation that consists of 1.7 million grid points. One time step of this simulation is calculated on eight nodes of the SP1 in the same time as required by a Cray Y/MP supercomputer. For the same simulation, 32-nodes of the SP1 and SP2 are required to reach the performance of a Cray C-90. A 32 node SP1 (SP2) configuration is 2.9 (4.6) times faster than a Cray Y/MP for this simulation, while the hypercube is roughly 2 times slower than the Y/MP for this application. KEY WORDS: Spatial direct numerical simulations; incompressible viscous flows; spectral methods; finite differences; parallel computing.

Enhancing Scalability and Efficiency of the TOUGH2_MP for LinuxClusters

DOE Office of Scientific and Technical Information (OSTI.GOV)

Zhang, Keni; Wu, Yu-Shu

2006-04-17

TOUGH2{_}MP, the parallel version TOUGH2 code, has been enhanced by implementing more efficient communication schemes. This enhancement is achieved through reducing the amount of small-size messages and the volume of large messages. The message exchange speed is further improved by using non-blocking communications for both linear and nonlinear iterations. In addition, we have modified the AZTEC parallel linear-equation solver to nonblocking communication. Through the improvement of code structuring and bug fixing, the new version code is now more stable, while demonstrating similar or even better nonlinear iteration converging speed than the original TOUGH2 code. As a result, the new versionmore » of TOUGH2{_}MP is improved significantly in its efficiency. In this paper, the scalability and efficiency of the parallel code are demonstrated by solving two large-scale problems. The testing results indicate that speedup of the code may depend on both problem size and complexity. In general, the code has excellent scalability in memory requirement as well as computing time.« less

THC-MP: High performance numerical simulation of reactive transport and multiphase flow in porous media

NASA Astrophysics Data System (ADS)

Wei, Xiaohui; Li, Weishan; Tian, Hailong; Li, Hongliang; Xu, Haixiao; Xu, Tianfu

2015-07-01

The numerical simulation of multiphase flow and reactive transport in the porous media on complex subsurface problem is a computationally intensive application. To meet the increasingly computational requirements, this paper presents a parallel computing method and architecture. Derived from TOUGHREACT that is a well-established code for simulating subsurface multi-phase flow and reactive transport problems, we developed a high performance computing THC-MP based on massive parallel computer, which extends greatly on the computational capability for the original code. The domain decomposition method was applied to the coupled numerical computing procedure in the THC-MP. We designed the distributed data structure, implemented the data initialization and exchange between the computing nodes and the core solving module using the hybrid parallel iterative and direct solver. Numerical accuracy of the THC-MP was verified through a CO2 injection-induced reactive transport problem by comparing the results obtained from the parallel computing and sequential computing (original code). Execution efficiency and code scalability were examined through field scale carbon sequestration applications on the multicore cluster. The results demonstrate successfully the enhanced performance using the THC-MP on parallel computing facilities.

A Numerical Study of Scalable Cardiac Electro-Mechanical Solvers on HPC Architectures

PubMed Central

Colli Franzone, Piero; Pavarino, Luca F.; Scacchi, Simone

2018-01-01

We introduce and study some scalable domain decomposition preconditioners for cardiac electro-mechanical 3D simulations on parallel HPC (High Performance Computing) architectures. The electro-mechanical model of the cardiac tissue is composed of four coupled sub-models: (1) the static finite elasticity equations for the transversely isotropic deformation of the cardiac tissue; (2) the active tension model describing the dynamics of the intracellular calcium, cross-bridge binding and myofilament tension; (3) the anisotropic Bidomain model describing the evolution of the intra- and extra-cellular potentials in the deforming cardiac tissue; and (4) the ionic membrane model describing the dynamics of ionic currents, gating variables, ionic concentrations and stretch-activated channels. This strongly coupled electro-mechanical model is discretized in time with a splitting semi-implicit technique and in space with isoparametric finite elements. The resulting scalable parallel solver is based on Multilevel Additive Schwarz preconditioners for the solution of the Bidomain system and on BDDC preconditioned Newton-Krylov solvers for the non-linear finite elasticity system. The results of several 3D parallel simulations show the scalability of both linear and non-linear solvers and their application to the study of both physiological excitation-contraction cardiac dynamics and re-entrant waves in the presence of different mechano-electrical feedbacks. PMID:29674971

Large-Scale Parallel Viscous Flow Computations using an Unstructured Multigrid Algorithm

NASA Technical Reports Server (NTRS)

Mavriplis, Dimitri J.

1999-01-01

The development and testing of a parallel unstructured agglomeration multigrid algorithm for steady-state aerodynamic flows is discussed. The agglomeration multigrid strategy uses a graph algorithm to construct the coarse multigrid levels from the given fine grid, similar to an algebraic multigrid approach, but operates directly on the non-linear system using the FAS (Full Approximation Scheme) approach. The scalability and convergence rate of the multigrid algorithm are examined on the SGI Origin 2000 and the Cray T3E. An argument is given which indicates that the asymptotic scalability of the multigrid algorithm should be similar to that of its underlying single grid smoothing scheme. For medium size problems involving several million grid points, near perfect scalability is obtained for the single grid algorithm, while only a slight drop-off in parallel efficiency is observed for the multigrid V- and W-cycles, using up to 128 processors on the SGI Origin 2000, and up to 512 processors on the Cray T3E. For a large problem using 25 million grid points, good scalability is observed for the multigrid algorithm using up to 1450 processors on a Cray T3E, even when the coarsest grid level contains fewer points than the total number of processors.

Computing Maximum Cardinality Matchings in Parallel on Bipartite Graphs via Tree-Grafting

DOE Office of Scientific and Technical Information (OSTI.GOV)

Azad, Ariful; Buluc, Aydn; Pothen, Alex

It is difficult to obtain high performance when computing matchings on parallel processors because matching algorithms explicitly or implicitly search for paths in the graph, and when these paths become long, there is little concurrency. In spite of this limitation, we present a new algorithm and its shared-memory parallelization that achieves good performance and scalability in computing maximum cardinality matchings in bipartite graphs. This algorithm searches for augmenting paths via specialized breadth-first searches (BFS) from multiple source vertices, hence creating more parallelism than single source algorithms. Algorithms that employ multiple-source searches cannot discard a search tree once no augmenting pathmore » is discovered from the tree, unlike algorithms that rely on single-source searches. We describe a novel tree-grafting method that eliminates most of the redundant edge traversals resulting from this property of multiple-source searches. We also employ the recent direction-optimizing BFS algorithm as a subroutine to discover augmenting paths faster. Our algorithm compares favorably with the current best algorithms in terms of the number of edges traversed, the average augmenting path length, and the number of iterations. Here, we provide a proof of correctness for our algorithm. Our NUMA-aware implementation is scalable to 80 threads of an Intel multiprocessor and to 240 threads on an Intel Knights Corner coprocessor. On average, our parallel algorithm runs an order of magnitude faster than the fastest algorithms available. The performance improvement is more significant on graphs with small matching number.« less

Computing Maximum Cardinality Matchings in Parallel on Bipartite Graphs via Tree-Grafting

DOE PAGES

Azad, Ariful; Buluc, Aydn; Pothen, Alex

2016-03-24

It is difficult to obtain high performance when computing matchings on parallel processors because matching algorithms explicitly or implicitly search for paths in the graph, and when these paths become long, there is little concurrency. In spite of this limitation, we present a new algorithm and its shared-memory parallelization that achieves good performance and scalability in computing maximum cardinality matchings in bipartite graphs. This algorithm searches for augmenting paths via specialized breadth-first searches (BFS) from multiple source vertices, hence creating more parallelism than single source algorithms. Algorithms that employ multiple-source searches cannot discard a search tree once no augmenting pathmore » is discovered from the tree, unlike algorithms that rely on single-source searches. We describe a novel tree-grafting method that eliminates most of the redundant edge traversals resulting from this property of multiple-source searches. We also employ the recent direction-optimizing BFS algorithm as a subroutine to discover augmenting paths faster. Our algorithm compares favorably with the current best algorithms in terms of the number of edges traversed, the average augmenting path length, and the number of iterations. Here, we provide a proof of correctness for our algorithm. Our NUMA-aware implementation is scalable to 80 threads of an Intel multiprocessor and to 240 threads on an Intel Knights Corner coprocessor. On average, our parallel algorithm runs an order of magnitude faster than the fastest algorithms available. The performance improvement is more significant on graphs with small matching number.« less

A distributed parallel storage architecture and its potential application within EOSDIS

NASA Technical Reports Server (NTRS)

Johnston, William E.; Tierney, Brian; Feuquay, Jay; Butzer, Tony

1994-01-01

We describe the architecture, implementation, use of a scalable, high performance, distributed-parallel data storage system developed in the ARPA funded MAGIC gigabit testbed. A collection of wide area distributed disk servers operate in parallel to provide logical block level access to large data sets. Operated primarily as a network-based cache, the architecture supports cooperation among independently owned resources to provide fast, large-scale, on-demand storage to support data handling, simulation, and computation.

OceanXtremes: Scalable Anomaly Detection in Oceanographic Time-Series

NASA Astrophysics Data System (ADS)

Wilson, B. D.; Armstrong, E. M.; Chin, T. M.; Gill, K. M.; Greguska, F. R., III; Huang, T.; Jacob, J. C.; Quach, N.

2016-12-01

The oceanographic community must meet the challenge to rapidly identify features and anomalies in complex and voluminous observations to further science and improve decision support. Given this data-intensive reality, we are developing an anomaly detection system, called OceanXtremes, powered by an intelligent, elastic Cloud-based analytic service backend that enables execution of domain-specific, multi-scale anomaly and feature detection algorithms across the entire archive of 15 to 30-year ocean science datasets.Our parallel analytics engine is extending the NEXUS system and exploits multiple open-source technologies: Apache Cassandra as a distributed spatial "tile" cache, Apache Spark for in-memory parallel computation, and Apache Solr for spatial search and storing pre-computed tile statistics and other metadata. OceanXtremes provides these key capabilities: Parallel generation (Spark on a compute cluster) of 15 to 30-year Ocean Climatologies (e.g. sea surface temperature or SST) in hours or overnight, using simple pixel averages or customizable Gaussian-weighted "smoothing" over latitude, longitude, and time; Parallel pre-computation, tiling, and caching of anomaly fields (daily variables minus a chosen climatology) with pre-computed tile statistics; Parallel detection (over the time-series of tiles) of anomalies or phenomena by regional area-averages exceeding a specified threshold (e.g. high SST in El Nino or SST "blob" regions), or more complex, custom data mining algorithms; Shared discovery and exploration of ocean phenomena and anomalies (facet search using Solr), along with unexpected correlations between key measured variables; Scalable execution for all capabilities on a hybrid Cloud, using our on-premise OpenStack Cloud cluster or at Amazon. The key idea is that the parallel data-mining operations will be run "near" the ocean data archives (a local "network" hop) so that we can efficiently access the thousands of files making up a three decade time-series. The presentation will cover the architecture of OceanXtremes, parallelization of the climatology computation and anomaly detection algorithms using Spark, example results for SST and other time-series, and parallel performance metrics.

A parallel computational model for GATE simulations.

PubMed

Rannou, F R; Vega-Acevedo, N; El Bitar, Z

2013-12-01

GATE/Geant4 Monte Carlo simulations are computationally demanding applications, requiring thousands of processor hours to produce realistic results. The classical strategy of distributing the simulation of individual events does not apply efficiently for Positron Emission Tomography (PET) experiments, because it requires a centralized coincidence processing and large communication overheads. We propose a parallel computational model for GATE that handles event generation and coincidence processing in a simple and efficient way by decentralizing event generation and processing but maintaining a centralized event and time coordinator. The model is implemented with the inclusion of a new set of factory classes that can run the same executable in sequential or parallel mode. A Mann-Whitney test shows that the output produced by this parallel model in terms of number of tallies is equivalent (but not equal) to its sequential counterpart. Computational performance evaluation shows that the software is scalable and well balanced. Copyright © 2013 Elsevier Ireland Ltd. All rights reserved.

A high performance linear equation solver on the VPP500 parallel supercomputer

DOE Office of Scientific and Technical Information (OSTI.GOV)

Nakanishi, Makoto; Ina, Hiroshi; Miura, Kenichi

1994-12-31

This paper describes the implementation of two high performance linear equation solvers developed for the Fujitsu VPP500, a distributed memory parallel supercomputer system. The solvers take advantage of the key architectural features of VPP500--(1) scalability for an arbitrary number of processors up to 222 processors, (2) flexible data transfer among processors provided by a crossbar interconnection network, (3) vector processing capability on each processor, and (4) overlapped computation and transfer. The general linear equation solver based on the blocked LU decomposition method achieves 120.0 GFLOPS performance with 100 processors in the LIN-PACK Highly Parallel Computing benchmark.

MOLNs: A CLOUD PLATFORM FOR INTERACTIVE, REPRODUCIBLE, AND SCALABLE SPATIAL STOCHASTIC COMPUTATIONAL EXPERIMENTS IN SYSTEMS BIOLOGY USING PyURDME.

PubMed

Drawert, Brian; Trogdon, Michael; Toor, Salman; Petzold, Linda; Hellander, Andreas

2016-01-01

Computational experiments using spatial stochastic simulations have led to important new biological insights, but they require specialized tools and a complex software stack, as well as large and scalable compute and data analysis resources due to the large computational cost associated with Monte Carlo computational workflows. The complexity of setting up and managing a large-scale distributed computation environment to support productive and reproducible modeling can be prohibitive for practitioners in systems biology. This results in a barrier to the adoption of spatial stochastic simulation tools, effectively limiting the type of biological questions addressed by quantitative modeling. In this paper, we present PyURDME, a new, user-friendly spatial modeling and simulation package, and MOLNs, a cloud computing appliance for distributed simulation of stochastic reaction-diffusion models. MOLNs is based on IPython and provides an interactive programming platform for development of sharable and reproducible distributed parallel computational experiments.

The novel high-performance 3-D MT inverse solver

NASA Astrophysics Data System (ADS)

Kruglyakov, Mikhail; Geraskin, Alexey; Kuvshinov, Alexey

2016-04-01

We present novel, robust, scalable, and fast 3-D magnetotelluric (MT) inverse solver. The solver is written in multi-language paradigm to make it as efficient, readable and maintainable as possible. Separation of concerns and single responsibility concepts go through implementation of the solver. As a forward modelling engine a modern scalable solver extrEMe, based on contracting integral equation approach, is used. Iterative gradient-type (quasi-Newton) optimization scheme is invoked to search for (regularized) inverse problem solution, and adjoint source approach is used to calculate efficiently the gradient of the misfit. The inverse solver is able to deal with highly detailed and contrasting models, allows for working (separately or jointly) with any type of MT responses, and supports massive parallelization. Moreover, different parallelization strategies implemented in the code allow optimal usage of available computational resources for a given problem statement. To parameterize an inverse domain the so-called mask parameterization is implemented, which means that one can merge any subset of forward modelling cells in order to account for (usually) irregular distribution of observation sites. We report results of 3-D numerical experiments aimed at analysing the robustness, performance and scalability of the code. In particular, our computational experiments carried out at different platforms ranging from modern laptops to HPC Piz Daint (6th supercomputer in the world) demonstrate practically linear scalability of the code up to thousands of nodes.

Cloud Computing Boosts Business Intelligence of Telecommunication Industry

NASA Astrophysics Data System (ADS)

Xu, Meng; Gao, Dan; Deng, Chao; Luo, Zhiguo; Sun, Shaoling

Business Intelligence becomes an attracting topic in today's data intensive applications, especially in telecommunication industry. Meanwhile, Cloud Computing providing IT supporting Infrastructure with excellent scalability, large scale storage, and high performance becomes an effective way to implement parallel data processing and data mining algorithms. BC-PDM (Big Cloud based Parallel Data Miner) is a new MapReduce based parallel data mining platform developed by CMRI (China Mobile Research Institute) to fit the urgent requirements of business intelligence in telecommunication industry. In this paper, the architecture, functionality and performance of BC-PDM are presented, together with the experimental evaluation and case studies of its applications. The evaluation result demonstrates both the usability and the cost-effectiveness of Cloud Computing based Business Intelligence system in applications of telecommunication industry.

SeqPig: simple and scalable scripting for large sequencing data sets in Hadoop.

PubMed

Schumacher, André; Pireddu, Luca; Niemenmaa, Matti; Kallio, Aleksi; Korpelainen, Eija; Zanetti, Gianluigi; Heljanko, Keijo

2014-01-01

Hadoop MapReduce-based approaches have become increasingly popular due to their scalability in processing large sequencing datasets. However, as these methods typically require in-depth expertise in Hadoop and Java, they are still out of reach of many bioinformaticians. To solve this problem, we have created SeqPig, a library and a collection of tools to manipulate, analyze and query sequencing datasets in a scalable and simple manner. SeqPigscripts use the Hadoop-based distributed scripting engine Apache Pig, which automatically parallelizes and distributes data processing tasks. We demonstrate SeqPig's scalability over many computing nodes and illustrate its use with example scripts. Available under the open source MIT license at http://sourceforge.net/projects/seqpig/

Scalable Implementation of Finite Elements by NASA _ Implicit (ScIFEi)

NASA Technical Reports Server (NTRS)

Warner, James E.; Bomarito, Geoffrey F.; Heber, Gerd; Hochhalter, Jacob D.

2016-01-01

Scalable Implementation of Finite Elements by NASA (ScIFEN) is a parallel finite element analysis code written in C++. ScIFEN is designed to provide scalable solutions to computational mechanics problems. It supports a variety of finite element types, nonlinear material models, and boundary conditions. This report provides an overview of ScIFEi (\\Sci-Fi"), the implicit solid mechanics driver within ScIFEN. A description of ScIFEi's capabilities is provided, including an overview of the tools and features that accompany the software as well as a description of the input and output le formats. Results from several problems are included, demonstrating the efficiency and scalability of ScIFEi by comparing to finite element analysis using a commercial code.

Parallel processing architecture for H.264 deblocking filter on multi-core platforms

NASA Astrophysics Data System (ADS)

Prasad, Durga P.; Sonachalam, Sekar; Kunchamwar, Mangesh K.; Gunupudi, Nageswara Rao

2012-03-01

Massively parallel computing (multi-core) chips offer outstanding new solutions that satisfy the increasing demand for high resolution and high quality video compression technologies such as H.264. Such solutions not only provide exceptional quality but also efficiency, low power, and low latency, previously unattainable in software based designs. While custom hardware and Application Specific Integrated Circuit (ASIC) technologies may achieve lowlatency, low power, and real-time performance in some consumer devices, many applications require a flexible and scalable software-defined solution. The deblocking filter in H.264 encoder/decoder poses difficult implementation challenges because of heavy data dependencies and the conditional nature of the computations. Deblocking filter implementations tend to be fixed and difficult to reconfigure for different needs. The ability to scale up for higher quality requirements such as 10-bit pixel depth or a 4:2:2 chroma format often reduces the throughput of a parallel architecture designed for lower feature set. A scalable architecture for deblocking filtering, created with a massively parallel processor based solution, means that the same encoder or decoder will be deployed in a variety of applications, at different video resolutions, for different power requirements, and at higher bit-depths and better color sub sampling patterns like YUV, 4:2:2, or 4:4:4 formats. Low power, software-defined encoders/decoders may be implemented using a massively parallel processor array, like that found in HyperX technology, with 100 or more cores and distributed memory. The large number of processor elements allows the silicon device to operate more efficiently than conventional DSP or CPU technology. This software programing model for massively parallel processors offers a flexible implementation and a power efficiency close to that of ASIC solutions. This work describes a scalable parallel architecture for an H.264 compliant deblocking filter for multi core platforms such as HyperX technology. Parallel techniques such as parallel processing of independent macroblocks, sub blocks, and pixel row level are examined in this work. The deblocking architecture consists of a basic cell called deblocking filter unit (DFU) and dependent data buffer manager (DFM). The DFU can be used in several instances, catering to different performance needs the DFM serves the data required for the different number of DFUs, and also manages all the neighboring data required for future data processing of DFUs. This approach achieves the scalability, flexibility, and performance excellence required in deblocking filters.

TESS

DOE Office of Scientific and Technical Information (OSTI.GOV)

Dmitriy Morozov, Tom Peterka

2014-07-29

Computing a Voronoi or Delaunay tessellation from a set of points is a core part of the analysis of many simulated and measured datasets. As the scale of simulations and observations surpasses billions of particles, a distributed-memory scalable parallel algorithm is the only feasible approach. The primary contribution of this software is a distributed-memory parallel Delaunay and Voronoi tessellation algorithm based on existing serial computational geometry libraries that automatically determines which neighbor points need to be exchanged among the subdomains of a spatial decomposition. Other contributions include the addition of periodic and wall boundary conditions.

A Scalable Software Architecture Booting and Configuring Nodes in the Whitney Commodity Computing Testbed

NASA Technical Reports Server (NTRS)

Fineberg, Samuel A.; Kutler, Paul (Technical Monitor)

1997-01-01

The Whitney project is integrating commodity off-the-shelf PC hardware and software technology to build a parallel supercomputer with hundreds to thousands of nodes. To build such a system, one must have a scalable software model, and the installation and maintenance of the system software must be completely automated. We describe the design of an architecture for booting, installing, and configuring nodes in such a system with particular consideration given to scalability and ease of maintenance. This system has been implemented on a 40-node prototype of Whitney and is to be used on the 500 processor Whitney system to be built in 1998.

Parallel computing in experimental mechanics and optical measurement: A review (II)

NASA Astrophysics Data System (ADS)

Wang, Tianyi; Kemao, Qian

2018-05-01

With advantages such as non-destructiveness, high sensitivity and high accuracy, optical techniques have successfully integrated into various important physical quantities in experimental mechanics (EM) and optical measurement (OM). However, in pursuit of higher image resolutions for higher accuracy, the computation burden of optical techniques has become much heavier. Therefore, in recent years, heterogeneous platforms composing of hardware such as CPUs and GPUs, have been widely employed to accelerate these techniques due to their cost-effectiveness, short development cycle, easy portability, and high scalability. In this paper, we analyze various works by first illustrating their different architectures, followed by introducing their various parallel patterns for high speed computation. Next, we review the effects of CPU and GPU parallel computing specifically in EM & OM applications in a broad scope, which include digital image/volume correlation, fringe pattern analysis, tomography, hyperspectral imaging, computer-generated holograms, and integral imaging. In our survey, we have found that high parallelism can always be exploited in such applications for the development of high-performance systems.

Efficient Parallelization of a Dynamic Unstructured Application on the Tera MTA

NASA Technical Reports Server (NTRS)

Oliker, Leonid; Biswas, Rupak

1999-01-01

The success of parallel computing in solving real-life computationally-intensive problems relies on their efficient mapping and execution on large-scale multiprocessor architectures. Many important applications are both unstructured and dynamic in nature, making their efficient parallel implementation a daunting task. This paper presents the parallelization of a dynamic unstructured mesh adaptation algorithm using three popular programming paradigms on three leading supercomputers. We examine an MPI message-passing implementation on the Cray T3E and the SGI Origin2OOO, a shared-memory implementation using cache coherent nonuniform memory access (CC-NUMA) of the Origin2OOO, and a multi-threaded version on the newly-released Tera Multi-threaded Architecture (MTA). We compare several critical factors of this parallel code development, including runtime, scalability, programmability, and memory overhead. Our overall results demonstrate that multi-threaded systems offer tremendous potential for quickly and efficiently solving some of the most challenging real-life problems on parallel computers.

Equalizer: a scalable parallel rendering framework.

PubMed

Eilemann, Stefan; Makhinya, Maxim; Pajarola, Renato

2009-01-01

Continuing improvements in CPU and GPU performances as well as increasing multi-core processor and cluster-based parallelism demand for flexible and scalable parallel rendering solutions that can exploit multipipe hardware accelerated graphics. In fact, to achieve interactive visualization, scalable rendering systems are essential to cope with the rapid growth of data sets. However, parallel rendering systems are non-trivial to develop and often only application specific implementations have been proposed. The task of developing a scalable parallel rendering framework is even more difficult if it should be generic to support various types of data and visualization applications, and at the same time work efficiently on a cluster with distributed graphics cards. In this paper we introduce a novel system called Equalizer, a toolkit for scalable parallel rendering based on OpenGL which provides an application programming interface (API) to develop scalable graphics applications for a wide range of systems ranging from large distributed visualization clusters and multi-processor multipipe graphics systems to single-processor single-pipe desktop machines. We describe the system architecture, the basic API, discuss its advantages over previous approaches, present example configurations and usage scenarios as well as scalability results.

Parallelized multi–graphics processing unit framework for high-speed Gabor-domain optical coherence microscopy

PubMed Central

Tankam, Patrice; Santhanam, Anand P.; Lee, Kye-Sung; Won, Jungeun; Canavesi, Cristina; Rolland, Jannick P.

2014-01-01

Abstract. Gabor-domain optical coherence microscopy (GD-OCM) is a volumetric high-resolution technique capable of acquiring three-dimensional (3-D) skin images with histological resolution. Real-time image processing is needed to enable GD-OCM imaging in a clinical setting. We present a parallelized and scalable multi-graphics processing unit (GPU) computing framework for real-time GD-OCM image processing. A parallelized control mechanism was developed to individually assign computation tasks to each of the GPUs. For each GPU, the optimal number of amplitude-scans (A-scans) to be processed in parallel was selected to maximize GPU memory usage and core throughput. We investigated five computing architectures for computational speed-up in processing 1000×1000 A-scans. The proposed parallelized multi-GPU computing framework enables processing at a computational speed faster than the GD-OCM image acquisition, thereby facilitating high-speed GD-OCM imaging in a clinical setting. Using two parallelized GPUs, the image processing of a 1×1×0.6  mm3 skin sample was performed in about 13 s, and the performance was benchmarked at 6.5 s with four GPUs. This work thus demonstrates that 3-D GD-OCM data may be displayed in real-time to the examiner using parallelized GPU processing. PMID:24695868

Parallelized multi-graphics processing unit framework for high-speed Gabor-domain optical coherence microscopy.

PubMed

Tankam, Patrice; Santhanam, Anand P; Lee, Kye-Sung; Won, Jungeun; Canavesi, Cristina; Rolland, Jannick P

2014-07-01

Gabor-domain optical coherence microscopy (GD-OCM) is a volumetric high-resolution technique capable of acquiring three-dimensional (3-D) skin images with histological resolution. Real-time image processing is needed to enable GD-OCM imaging in a clinical setting. We present a parallelized and scalable multi-graphics processing unit (GPU) computing framework for real-time GD-OCM image processing. A parallelized control mechanism was developed to individually assign computation tasks to each of the GPUs. For each GPU, the optimal number of amplitude-scans (A-scans) to be processed in parallel was selected to maximize GPU memory usage and core throughput. We investigated five computing architectures for computational speed-up in processing 1000×1000 A-scans. The proposed parallelized multi-GPU computing framework enables processing at a computational speed faster than the GD-OCM image acquisition, thereby facilitating high-speed GD-OCM imaging in a clinical setting. Using two parallelized GPUs, the image processing of a 1×1×0.6  mm3 skin sample was performed in about 13 s, and the performance was benchmarked at 6.5 s with four GPUs. This work thus demonstrates that 3-D GD-OCM data may be displayed in real-time to the examiner using parallelized GPU processing.

PARALLEL HOP: A SCALABLE HALO FINDER FOR MASSIVE COSMOLOGICAL DATA SETS

DOE Office of Scientific and Technical Information (OSTI.GOV)

Skory, Stephen; Turk, Matthew J.; Norman, Michael L.

2010-11-15

Modern N-body cosmological simulations contain billions (10{sup 9}) of dark matter particles. These simulations require hundreds to thousands of gigabytes of memory and employ hundreds to tens of thousands of processing cores on many compute nodes. In order to study the distribution of dark matter in a cosmological simulation, the dark matter halos must be identified using a halo finder, which establishes the halo membership of every particle in the simulation. The resources required for halo finding are similar to the requirements for the simulation itself. In particular, simulations have become too extensive to use commonly employed halo finders, suchmore » that the computational requirements to identify halos must now be spread across multiple nodes and cores. Here, we present a scalable-parallel halo finding method called Parallel HOP for large-scale cosmological simulation data. Based on the halo finder HOP, it utilizes message passing interface and domain decomposition to distribute the halo finding workload across multiple compute nodes, enabling analysis of much larger data sets than is possible with the strictly serial or previous parallel implementations of HOP. We provide a reference implementation of this method as a part of the toolkit {sup yt}, an analysis toolkit for adaptive mesh refinement data that include complementary analysis modules. Additionally, we discuss a suite of benchmarks that demonstrate that this method scales well up to several hundred tasks and data sets in excess of 2000{sup 3} particles. The Parallel HOP method and our implementation can be readily applied to any kind of N-body simulation data and is therefore widely applicable.« less

Scalable Algorithms for Clustering Large Geospatiotemporal Data Sets on Manycore Architectures

NASA Astrophysics Data System (ADS)

Mills, R. T.; Hoffman, F. M.; Kumar, J.; Sreepathi, S.; Sripathi, V.

2016-12-01

The increasing availability of high-resolution geospatiotemporal data sets from sources such as observatory networks, remote sensing platforms, and computational Earth system models has opened new possibilities for knowledge discovery using data sets fused from disparate sources. Traditional algorithms and computing platforms are impractical for the analysis and synthesis of data sets of this size; however, new algorithmic approaches that can effectively utilize the complex memory hierarchies and the extremely high levels of available parallelism in state-of-the-art high-performance computing platforms can enable such analysis. We describe a massively parallel implementation of accelerated k-means clustering and some optimizations to boost computational intensity and utilization of wide SIMD lanes on state-of-the art multi- and manycore processors, including the second-generation Intel Xeon Phi ("Knights Landing") processor based on the Intel Many Integrated Core (MIC) architecture, which includes several new features, including an on-package high-bandwidth memory. We also analyze the code in the context of a few practical applications to the analysis of climatic and remotely-sensed vegetation phenology data sets, and speculate on some of the new applications that such scalable analysis methods may enable.

An efficient implementation of a high-order filter for a cubed-sphere spectral element model

NASA Astrophysics Data System (ADS)

Kang, Hyun-Gyu; Cheong, Hyeong-Bin

2017-03-01

A parallel-scalable, isotropic, scale-selective spatial filter was developed for the cubed-sphere spectral element model on the sphere. The filter equation is a high-order elliptic (Helmholtz) equation based on the spherical Laplacian operator, which is transformed into cubed-sphere local coordinates. The Laplacian operator is discretized on the computational domain, i.e., on each cell, by the spectral element method with Gauss-Lobatto Lagrange interpolating polynomials (GLLIPs) as the orthogonal basis functions. On the global domain, the discrete filter equation yielded a linear system represented by a highly sparse matrix. The density of this matrix increases quadratically (linearly) with the order of GLLIP (order of the filter), and the linear system is solved in only O (Ng) operations, where Ng is the total number of grid points. The solution, obtained by a row reduction method, demonstrated the typical accuracy and convergence rate of the cubed-sphere spectral element method. To achieve computational efficiency on parallel computers, the linear system was treated by an inverse matrix method (a sparse matrix-vector multiplication). The density of the inverse matrix was lowered to only a few times of the original sparse matrix without degrading the accuracy of the solution. For better computational efficiency, a local-domain high-order filter was introduced: The filter equation is applied to multiple cells, and then the central cell was only used to reconstruct the filtered field. The parallel efficiency of applying the inverse matrix method to the global- and local-domain filter was evaluated by the scalability on a distributed-memory parallel computer. The scale-selective performance of the filter was demonstrated on Earth topography. The usefulness of the filter as a hyper-viscosity for the vorticity equation was also demonstrated.

Large-scale molecular dynamics simulation of DNA: implementation and validation of the AMBER98 force field in LAMMPS.

PubMed

Grindon, Christina; Harris, Sarah; Evans, Tom; Novik, Keir; Coveney, Peter; Laughton, Charles

2004-07-15

Molecular modelling played a central role in the discovery of the structure of DNA by Watson and Crick. Today, such modelling is done on computers: the more powerful these computers are, the more detailed and extensive can be the study of the dynamics of such biological macromolecules. To fully harness the power of modern massively parallel computers, however, we need to develop and deploy algorithms which can exploit the structure of such hardware. The Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) is a scalable molecular dynamics code including long-range Coulomb interactions, which has been specifically designed to function efficiently on parallel platforms. Here we describe the implementation of the AMBER98 force field in LAMMPS and its validation for molecular dynamics investigations of DNA structure and flexibility against the benchmark of results obtained with the long-established code AMBER6 (Assisted Model Building with Energy Refinement, version 6). Extended molecular dynamics simulations on the hydrated DNA dodecamer d(CTTTTGCAAAAG)(2), which has previously been the subject of extensive dynamical analysis using AMBER6, show that it is possible to obtain excellent agreement in terms of static, dynamic and thermodynamic parameters between AMBER6 and LAMMPS. In comparison with AMBER6, LAMMPS shows greatly improved scalability in massively parallel environments, opening up the possibility of efficient simulations of order-of-magnitude larger systems and/or for order-of-magnitude greater simulation times.

High Performance Parallel Computational Nanotechnology

NASA Technical Reports Server (NTRS)

Saini, Subhash; Craw, James M. (Technical Monitor)

1995-01-01

At a recent press conference, NASA Administrator Dan Goldin encouraged NASA Ames Research Center to take a lead role in promoting research and development of advanced, high-performance computer technology, including nanotechnology. Manufacturers of leading-edge microprocessors currently perform large-scale simulations in the design and verification of semiconductor devices and microprocessors. Recently, the need for this intensive simulation and modeling analysis has greatly increased, due in part to the ever-increasing complexity of these devices, as well as the lessons of experiences such as the Pentium fiasco. Simulation, modeling, testing, and validation will be even more important for designing molecular computers because of the complex specification of millions of atoms, thousands of assembly steps, as well as the simulation and modeling needed to ensure reliable, robust and efficient fabrication of the molecular devices. The software for this capacity does not exist today, but it can be extrapolated from the software currently used in molecular modeling for other applications: semi-empirical methods, ab initio methods, self-consistent field methods, Hartree-Fock methods, molecular mechanics; and simulation methods for diamondoid structures. In as much as it seems clear that the application of such methods in nanotechnology will require powerful, highly powerful systems, this talk will discuss techniques and issues for performing these types of computations on parallel systems. We will describe system design issues (memory, I/O, mass storage, operating system requirements, special user interface issues, interconnects, bandwidths, and programming languages) involved in parallel methods for scalable classical, semiclassical, quantum, molecular mechanics, and continuum models; molecular nanotechnology computer-aided designs (NanoCAD) techniques; visualization using virtual reality techniques of structural models and assembly sequences; software required to control mini robotic manipulators for positional control; scalable numerical algorithms for reliability, verifications and testability. There appears no fundamental obstacle to simulating molecular compilers and molecular computers on high performance parallel computers, just as the Boeing 777 was simulated on a computer before manufacturing it.

A high performance parallel algorithm for 1-D FFT

DOE Office of Scientific and Technical Information (OSTI.GOV)

Agarwal, R.C.; Gustavson, F.G.; Zubair, M.

1994-12-31

In this paper the authors propose a parallel high performance FFT algorithm based on a multi-dimensional formulation. They use this to solve a commonly encountered FFT based kernel on a distributed memory parallel machine, the IBM scalable parallel system, SP1. The kernel requires a forward FFT computation of an input sequence, multiplication of the transformed data by a coefficient array, and finally an inverse FFT computation of the resultant data. They show that the multi-dimensional formulation helps in reducing the communication costs and also improves the single node performance by effectively utilizing the memory system of the node. They implementedmore » this kernel on the IBM SP1 and observed a performance of 1.25 GFLOPS on a 64-node machine.« less

Spherical harmonic results for the 3D Kobayashi Benchmark suite

DOE Office of Scientific and Technical Information (OSTI.GOV)

Brown, P N; Chang, B; Hanebutte, U R

1999-03-02

Spherical harmonic solutions are presented for the Kobayashi benchmark suite. The results were obtained with Ardra, a scalable, parallel neutron transport code developed at Lawrence Livermore National Laboratory (LLNL). The calculations were performed on the IBM ASCI Blue-Pacific computer at LLNL.

A parallel Monte Carlo code for planar and SPECT imaging: implementation, verification and applications in (131)I SPECT.

PubMed

Dewaraja, Yuni K; Ljungberg, Michael; Majumdar, Amitava; Bose, Abhijit; Koral, Kenneth F

2002-02-01

This paper reports the implementation of the SIMIND Monte Carlo code on an IBM SP2 distributed memory parallel computer. Basic aspects of running Monte Carlo particle transport calculations on parallel architectures are described. Our parallelization is based on equally partitioning photons among the processors and uses the Message Passing Interface (MPI) library for interprocessor communication and the Scalable Parallel Random Number Generator (SPRNG) to generate uncorrelated random number streams. These parallelization techniques are also applicable to other distributed memory architectures. A linear increase in computing speed with the number of processors is demonstrated for up to 32 processors. This speed-up is especially significant in Single Photon Emission Computed Tomography (SPECT) simulations involving higher energy photon emitters, where explicit modeling of the phantom and collimator is required. For (131)I, the accuracy of the parallel code is demonstrated by comparing simulated and experimental SPECT images from a heart/thorax phantom. Clinically realistic SPECT simulations using the voxel-man phantom are carried out to assess scatter and attenuation correction.

Time-dependent density-functional theory in massively parallel computer architectures: the octopus project

NASA Astrophysics Data System (ADS)

Andrade, Xavier; Alberdi-Rodriguez, Joseba; Strubbe, David A.; Oliveira, Micael J. T.; Nogueira, Fernando; Castro, Alberto; Muguerza, Javier; Arruabarrena, Agustin; Louie, Steven G.; Aspuru-Guzik, Alán; Rubio, Angel; Marques, Miguel A. L.

2012-06-01

Octopus is a general-purpose density-functional theory (DFT) code, with a particular emphasis on the time-dependent version of DFT (TDDFT). In this paper we present the ongoing efforts to achieve the parallelization of octopus. We focus on the real-time variant of TDDFT, where the time-dependent Kohn-Sham equations are directly propagated in time. This approach has great potential for execution in massively parallel systems such as modern supercomputers with thousands of processors and graphics processing units (GPUs). For harvesting the potential of conventional supercomputers, the main strategy is a multi-level parallelization scheme that combines the inherent scalability of real-time TDDFT with a real-space grid domain-partitioning approach. A scalable Poisson solver is critical for the efficiency of this scheme. For GPUs, we show how using blocks of Kohn-Sham states provides the required level of data parallelism and that this strategy is also applicable for code optimization on standard processors. Our results show that real-time TDDFT, as implemented in octopus, can be the method of choice for studying the excited states of large molecular systems in modern parallel architectures.

Time-dependent density-functional theory in massively parallel computer architectures: the OCTOPUS project.

PubMed

Andrade, Xavier; Alberdi-Rodriguez, Joseba; Strubbe, David A; Oliveira, Micael J T; Nogueira, Fernando; Castro, Alberto; Muguerza, Javier; Arruabarrena, Agustin; Louie, Steven G; Aspuru-Guzik, Alán; Rubio, Angel; Marques, Miguel A L

2012-06-13

Octopus is a general-purpose density-functional theory (DFT) code, with a particular emphasis on the time-dependent version of DFT (TDDFT). In this paper we present the ongoing efforts to achieve the parallelization of octopus. We focus on the real-time variant of TDDFT, where the time-dependent Kohn-Sham equations are directly propagated in time. This approach has great potential for execution in massively parallel systems such as modern supercomputers with thousands of processors and graphics processing units (GPUs). For harvesting the potential of conventional supercomputers, the main strategy is a multi-level parallelization scheme that combines the inherent scalability of real-time TDDFT with a real-space grid domain-partitioning approach. A scalable Poisson solver is critical for the efficiency of this scheme. For GPUs, we show how using blocks of Kohn-Sham states provides the required level of data parallelism and that this strategy is also applicable for code optimization on standard processors. Our results show that real-time TDDFT, as implemented in octopus, can be the method of choice for studying the excited states of large molecular systems in modern parallel architectures.

Performance-scalable volumetric data classification for online industrial inspection

NASA Astrophysics Data System (ADS)

Abraham, Aby J.; Sadki, Mustapha; Lea, R. M.

2002-03-01

Non-intrusive inspection and non-destructive testing of manufactured objects with complex internal structures typically requires the enhancement, analysis and visualization of high-resolution volumetric data. Given the increasing availability of fast 3D scanning technology (e.g. cone-beam CT), enabling on-line detection and accurate discrimination of components or sub-structures, the inherent complexity of classification algorithms inevitably leads to throughput bottlenecks. Indeed, whereas typical inspection throughput requirements range from 1 to 1000 volumes per hour, depending on density and resolution, current computational capability is one to two orders-of-magnitude less. Accordingly, speeding up classification algorithms requires both reduction of algorithm complexity and acceleration of computer performance. A shape-based classification algorithm, offering algorithm complexity reduction, by using ellipses as generic descriptors of solids-of-revolution, and supporting performance-scalability, by exploiting the inherent parallelism of volumetric data, is presented. A two-stage variant of the classical Hough transform is used for ellipse detection and correlation of the detected ellipses facilitates position-, scale- and orientation-invariant component classification. Performance-scalability is achieved cost-effectively by accelerating a PC host with one or more COTS (Commercial-Off-The-Shelf) PCI multiprocessor cards. Experimental results are reported to demonstrate the feasibility and cost-effectiveness of the data-parallel classification algorithm for on-line industrial inspection applications.

MOLNs: A CLOUD PLATFORM FOR INTERACTIVE, REPRODUCIBLE, AND SCALABLE SPATIAL STOCHASTIC COMPUTATIONAL EXPERIMENTS IN SYSTEMS BIOLOGY USING PyURDME

PubMed Central

Drawert, Brian; Trogdon, Michael; Toor, Salman; Petzold, Linda; Hellander, Andreas

2017-01-01

Computational experiments using spatial stochastic simulations have led to important new biological insights, but they require specialized tools and a complex software stack, as well as large and scalable compute and data analysis resources due to the large computational cost associated with Monte Carlo computational workflows. The complexity of setting up and managing a large-scale distributed computation environment to support productive and reproducible modeling can be prohibitive for practitioners in systems biology. This results in a barrier to the adoption of spatial stochastic simulation tools, effectively limiting the type of biological questions addressed by quantitative modeling. In this paper, we present PyURDME, a new, user-friendly spatial modeling and simulation package, and MOLNs, a cloud computing appliance for distributed simulation of stochastic reaction-diffusion models. MOLNs is based on IPython and provides an interactive programming platform for development of sharable and reproducible distributed parallel computational experiments. PMID:28190948

An Evaluation of Architectural Platforms for Parallel Navier-Stokes Computations

NASA Technical Reports Server (NTRS)

Jayasimha, D. N.; Hayder, M. E.; Pillay, S. K.

1996-01-01

We study the computational, communication, and scalability characteristics of a computational fluid dynamics application, which solves the time accurate flow field of a jet using the compressible Navier-Stokes equations, on a variety of parallel architecture platforms. The platforms chosen for this study are a cluster of workstations (the LACE experimental testbed at NASA Lewis), a shared memory multiprocessor (the Cray YMP), and distributed memory multiprocessors with different topologies - the IBM SP and the Cray T3D. We investigate the impact of various networks connecting the cluster of workstations on the performance of the application and the overheads induced by popular message passing libraries used for parallelization. The work also highlights the importance of matching the memory bandwidth to the processor speed for good single processor performance. By studying the performance of an application on a variety of architectures, we are able to point out the strengths and weaknesses of each of the example computing platforms.

Parallelizing Navier-Stokes Computations on a Variety of Architectural Platforms

NASA Technical Reports Server (NTRS)

Jayasimha, D. N.; Hayder, M. E.; Pillay, S. K.

1997-01-01

We study the computational, communication, and scalability characteristics of a Computational Fluid Dynamics application, which solves the time accurate flow field of a jet using the compressible Navier-Stokes equations, on a variety of parallel architectural platforms. The platforms chosen for this study are a cluster of workstations (the LACE experimental testbed at NASA Lewis), a shared memory multiprocessor (the Cray YMP), distributed memory multiprocessors with different topologies-the IBM SP and the Cray T3D. We investigate the impact of various networks, connecting the cluster of workstations, on the performance of the application and the overheads induced by popular message passing libraries used for parallelization. The work also highlights the importance of matching the memory bandwidth to the processor speed for good single processor performance. By studying the performance of an application on a variety of architectures, we are able to point out the strengths and weaknesses of each of the example computing platforms.

Execution of parallel algorithms on a heterogeneous multicomputer

NASA Astrophysics Data System (ADS)

Isenstein, Barry S.; Greene, Jonathon

1995-04-01

Many aerospace/defense sensing and dual-use applications require high-performance computing, extensive high-bandwidth interconnect and realtime deterministic operation. This paper will describe the architecture of a scalable multicomputer that includes DSP and RISC processors. A single chassis implementation is capable of delivering in excess of 10 GFLOPS of DSP processing power with 2 Gbytes/s of realtime sensor I/O. A software approach to implementing parallel algorithms called the Parallel Application System (PAS) is also presented. An example of applying PAS to a DSP application is shown.

Scalable Multiprocessor for High-Speed Computing in Space

NASA Technical Reports Server (NTRS)

Lux, James; Lang, Minh; Nishimoto, Kouji; Clark, Douglas; Stosic, Dorothy; Bachmann, Alex; Wilkinson, William; Steffke, Richard

2004-01-01

A report discusses the continuing development of a scalable multiprocessor computing system for hard real-time applications aboard a spacecraft. "Hard realtime applications" signifies applications, like real-time radar signal processing, in which the data to be processed are generated at "hundreds" of pulses per second, each pulse "requiring" millions of arithmetic operations. In these applications, the digital processors must be tightly integrated with analog instrumentation (e.g., radar equipment), and data input/output must be synchronized with analog instrumentation, controlled to within fractions of a microsecond. The scalable multiprocessor is a cluster of identical commercial-off-the-shelf generic DSP (digital-signal-processing) computers plus generic interface circuits, including analog-to-digital converters, all controlled by software. The processors are computers interconnected by high-speed serial links. Performance can be increased by adding hardware modules and correspondingly modifying the software. Work is distributed among the processors in a parallel or pipeline fashion by means of a flexible master/slave control and timing scheme. Each processor operates under its own local clock; synchronization is achieved by broadcasting master time signals to all the processors, which compute offsets between the master clock and their local clocks.

Visual Analytics for Power Grid Contingency Analysis

DOE Office of Scientific and Technical Information (OSTI.GOV)

Wong, Pak C.; Huang, Zhenyu; Chen, Yousu

2014-01-20

Contingency analysis is the process of employing different measures to model scenarios, analyze them, and then derive the best response to remove the threats. This application paper focuses on a class of contingency analysis problems found in the power grid management system. A power grid is a geographically distributed interconnected transmission network that transmits and delivers electricity from generators to end users. The power grid contingency analysis problem is increasingly important because of both the growing size of the underlying raw data that need to be analyzed and the urgency to deliver working solutions in an aggressive timeframe. Failure tomore » do so may bring significant financial, economic, and security impacts to all parties involved and the society at large. The paper presents a scalable visual analytics pipeline that transforms about 100 million contingency scenarios to a manageable size and form for grid operators to examine different scenarios and come up with preventive or mitigation strategies to address the problems in a predictive and timely manner. Great attention is given to the computational scalability, information scalability, visual scalability, and display scalability issues surrounding the data analytics pipeline. Most of the large-scale computation requirements of our work are conducted on a Cray XMT multi-threaded parallel computer. The paper demonstrates a number of examples using western North American power grid models and data.« less

Studying an Eulerian Computer Model on Different High-performance Computer Platforms and Some Applications

NASA Astrophysics Data System (ADS)

Georgiev, K.; Zlatev, Z.

2010-11-01

The Danish Eulerian Model (DEM) is an Eulerian model for studying the transport of air pollutants on large scale. Originally, the model was developed at the National Environmental Research Institute of Denmark. The model computational domain covers Europe and some neighbour parts belong to the Atlantic Ocean, Asia and Africa. If DEM model is to be applied by using fine grids, then its discretization leads to a huge computational problem. This implies that such a model as DEM must be run only on high-performance computer architectures. The implementation and tuning of such a complex large-scale model on each different computer is a non-trivial task. Here, some comparison results of running of this model on different kind of vector (CRAY C92A, Fujitsu, etc.), parallel computers with distributed memory (IBM SP, CRAY T3E, Beowulf clusters, Macintosh G4 clusters, etc.), parallel computers with shared memory (SGI Origin, SUN, etc.) and parallel computers with two levels of parallelism (IBM SMP, IBM BlueGene/P, clusters of multiprocessor nodes, etc.) will be presented. The main idea in the parallel version of DEM is domain partitioning approach. Discussions according to the effective use of the cache and hierarchical memories of the modern computers as well as the performance, speed-ups and efficiency achieved will be done. The parallel code of DEM, created by using MPI standard library, appears to be highly portable and shows good efficiency and scalability on different kind of vector and parallel computers. Some important applications of the computer model output are presented in short.

Scalable hierarchical PDE sampler for generating spatially correlated random fields using nonmatching meshes: Scalable hierarchical PDE sampler using nonmatching meshes

DOE PAGES

Osborn, Sarah; Zulian, Patrick; Benson, Thomas; ...

2018-01-30

This work describes a domain embedding technique between two nonmatching meshes used for generating realizations of spatially correlated random fields with applications to large-scale sampling-based uncertainty quantification. The goal is to apply the multilevel Monte Carlo (MLMC) method for the quantification of output uncertainties of PDEs with random input coefficients on general and unstructured computational domains. We propose a highly scalable, hierarchical sampling method to generate realizations of a Gaussian random field on a given unstructured mesh by solving a reaction–diffusion PDE with a stochastic right-hand side. The stochastic PDE is discretized using the mixed finite element method on anmore » embedded domain with a structured mesh, and then, the solution is projected onto the unstructured mesh. This work describes implementation details on how to efficiently transfer data from the structured and unstructured meshes at coarse levels, assuming that this can be done efficiently on the finest level. We investigate the efficiency and parallel scalability of the technique for the scalable generation of Gaussian random fields in three dimensions. An application of the MLMC method is presented for quantifying uncertainties of subsurface flow problems. Here, we demonstrate the scalability of the sampling method with nonmatching mesh embedding, coupled with a parallel forward model problem solver, for large-scale 3D MLMC simulations with up to 1.9·109 unknowns.« less

Scalable hierarchical PDE sampler for generating spatially correlated random fields using nonmatching meshes: Scalable hierarchical PDE sampler using nonmatching meshes

DOE Office of Scientific and Technical Information (OSTI.GOV)

Osborn, Sarah; Zulian, Patrick; Benson, Thomas

This work describes a domain embedding technique between two nonmatching meshes used for generating realizations of spatially correlated random fields with applications to large-scale sampling-based uncertainty quantification. The goal is to apply the multilevel Monte Carlo (MLMC) method for the quantification of output uncertainties of PDEs with random input coefficients on general and unstructured computational domains. We propose a highly scalable, hierarchical sampling method to generate realizations of a Gaussian random field on a given unstructured mesh by solving a reaction–diffusion PDE with a stochastic right-hand side. The stochastic PDE is discretized using the mixed finite element method on anmore » embedded domain with a structured mesh, and then, the solution is projected onto the unstructured mesh. This work describes implementation details on how to efficiently transfer data from the structured and unstructured meshes at coarse levels, assuming that this can be done efficiently on the finest level. We investigate the efficiency and parallel scalability of the technique for the scalable generation of Gaussian random fields in three dimensions. An application of the MLMC method is presented for quantifying uncertainties of subsurface flow problems. Here, we demonstrate the scalability of the sampling method with nonmatching mesh embedding, coupled with a parallel forward model problem solver, for large-scale 3D MLMC simulations with up to 1.9·109 unknowns.« less

Better than $l/Mflops sustained: a scalable PC-based parallel computer for lattice QCD

NASA Astrophysics Data System (ADS)

Fodor, Zoltán; Katz, Sándor D.; Papp, Gábor

2003-05-01

We study the feasibility of a PC-based parallel computer for medium to large scale lattice QCD simulations. The Eötvös Univ., Inst. Theor. Phys. cluster consists of 137 Intel P4-1.7GHz nodes with 512 MB RDRAM. The 32-bit, single precision sustained performance for dynamical QCD without communication is 1510 Mflops/node with Wilson and 970 Mflops/node with staggered fermions. This gives a total performance of 208 Gflops for Wilson and 133 Gflops for staggered QCD, respectively (for 64-bit applications the performance is approximately halved). The novel feature of our system is its communication architecture. In order to have a scalable, cost-effective machine we use Gigabit Ethernet cards for nearest-neighbor communications in a two-dimensional mesh. This type of communication is cost effective (only 30% of the hardware costs is spent on the communication). According to our benchmark measurements this type of communication results in around 40% communication time fraction for lattices upto 48 3·96 in full QCD simulations. The price/sustained-performance ratio for full QCD is better than l/Mflops for Wilson (and around 1.5/Mflops for staggered) quarks for practically any lattice size, which can fit in our parallel computer. The communication software is freely available upon request for non-profit organizations.

SeqPig: simple and scalable scripting for large sequencing data sets in Hadoop

PubMed Central

Schumacher, André; Pireddu, Luca; Niemenmaa, Matti; Kallio, Aleksi; Korpelainen, Eija; Zanetti, Gianluigi; Heljanko, Keijo

2014-01-01

Summary: Hadoop MapReduce-based approaches have become increasingly popular due to their scalability in processing large sequencing datasets. However, as these methods typically require in-depth expertise in Hadoop and Java, they are still out of reach of many bioinformaticians. To solve this problem, we have created SeqPig, a library and a collection of tools to manipulate, analyze and query sequencing datasets in a scalable and simple manner. SeqPigscripts use the Hadoop-based distributed scripting engine Apache Pig, which automatically parallelizes and distributes data processing tasks. We demonstrate SeqPig’s scalability over many computing nodes and illustrate its use with example scripts. Availability and Implementation: Available under the open source MIT license at http://sourceforge.net/projects/seqpig/ Contact: andre.schumacher@yahoo.com Supplementary information: Supplementary data are available at Bioinformatics online. PMID:24149054

Parallel O(N) Stokes’ solver towards scalable Brownian dynamics of hydrodynamically interacting objects in general geometries

DOE Office of Scientific and Technical Information (OSTI.GOV)

Zhao, Xujun; Li, Jiyuan; Jiang, Xikai

An efficient parallel Stokes’s solver is developed towards the complete inclusion of hydrodynamic interactions of Brownian particles in any geometry. A Langevin description of the particle dynamics is adopted, where the long-range interactions are included using a Green’s function formalism. We present a scalable parallel computational approach, where the general geometry Stokeslet is calculated following a matrix-free algorithm using the General geometry Ewald-like method. Our approach employs a highly-efficient iterative finite element Stokes’ solver for the accurate treatment of long-range hydrodynamic interactions within arbitrary confined geometries. A combination of mid-point time integration of the Brownian stochastic differential equation, the parallelmore » Stokes’ solver, and a Chebyshev polynomial approximation for the fluctuation-dissipation theorem result in an O(N) parallel algorithm. We also illustrate the new algorithm in the context of the dynamics of confined polymer solutions in equilibrium and non-equilibrium conditions. Our method is extended to treat suspended finite size particles of arbitrary shape in any geometry using an Immersed Boundary approach.« less

Parallel O(N) Stokes’ solver towards scalable Brownian dynamics of hydrodynamically interacting objects in general geometries

DOE PAGES

Zhao, Xujun; Li, Jiyuan; Jiang, Xikai; ...

2017-06-29

An efficient parallel Stokes’s solver is developed towards the complete inclusion of hydrodynamic interactions of Brownian particles in any geometry. A Langevin description of the particle dynamics is adopted, where the long-range interactions are included using a Green’s function formalism. We present a scalable parallel computational approach, where the general geometry Stokeslet is calculated following a matrix-free algorithm using the General geometry Ewald-like method. Our approach employs a highly-efficient iterative finite element Stokes’ solver for the accurate treatment of long-range hydrodynamic interactions within arbitrary confined geometries. A combination of mid-point time integration of the Brownian stochastic differential equation, the parallelmore » Stokes’ solver, and a Chebyshev polynomial approximation for the fluctuation-dissipation theorem result in an O(N) parallel algorithm. We also illustrate the new algorithm in the context of the dynamics of confined polymer solutions in equilibrium and non-equilibrium conditions. Our method is extended to treat suspended finite size particles of arbitrary shape in any geometry using an Immersed Boundary approach.« less

Parallel heuristics for scalable community detection

DOE PAGES

Lu, Hao; Halappanavar, Mahantesh; Kalyanaraman, Ananth

2015-08-14

Community detection has become a fundamental operation in numerous graph-theoretic applications. Despite its potential for application, there is only limited support for community detection on large-scale parallel computers, largely owing to the irregular and inherently sequential nature of the underlying heuristics. In this paper, we present parallelization heuristics for fast community detection using the Louvain method as the serial template. The Louvain method is an iterative heuristic for modularity optimization. Originally developed in 2008, the method has become increasingly popular owing to its ability to detect high modularity community partitions in a fast and memory-efficient manner. However, the method ismore » also inherently sequential, thereby limiting its scalability. Here, we observe certain key properties of this method that present challenges for its parallelization, and consequently propose heuristics that are designed to break the sequential barrier. For evaluation purposes, we implemented our heuristics using OpenMP multithreading, and tested them over real world graphs derived from multiple application domains. Compared to the serial Louvain implementation, our parallel implementation is able to produce community outputs with a higher modularity for most of the inputs tested, in comparable number or fewer iterations, while providing real speedups of up to 16x using 32 threads.« less

Automatic Parallelization of Numerical Python Applications using the Global Arrays Toolkit

DOE Office of Scientific and Technical Information (OSTI.GOV)

Daily, Jeffrey A.; Lewis, Robert R.

2011-11-30

Global Arrays is a software system from Pacific Northwest National Laboratory that enables an efficient, portable, and parallel shared-memory programming interface to manipulate distributed dense arrays. The NumPy module is the de facto standard for numerical calculation in the Python programming language, a language whose use is growing rapidly in the scientific and engineering communities. NumPy provides a powerful N-dimensional array class as well as other scientific computing capabilities. However, like the majority of the core Python modules, NumPy is inherently serial. Using a combination of Global Arrays and NumPy, we have reimplemented NumPy as a distributed drop-in replacement calledmore » Global Arrays in NumPy (GAiN). Serial NumPy applications can become parallel, scalable GAiN applications with only minor source code changes. Scalability studies of several different GAiN applications will be presented showing the utility of developing serial NumPy codes which can later run on more capable clusters or supercomputers.« less

A Massively Parallel Computational Method of Reading Index Files for SOAPsnv.

PubMed

Zhu, Xiaoqian; Peng, Shaoliang; Liu, Shaojie; Cui, Yingbo; Gu, Xiang; Gao, Ming; Fang, Lin; Fang, Xiaodong

2015-12-01

SOAPsnv is the software used for identifying the single nucleotide variation in cancer genes. However, its performance is yet to match the massive amount of data to be processed. Experiments reveal that the main performance bottleneck of SOAPsnv software is the pileup algorithm. The original pileup algorithm's I/O process is time-consuming and inefficient to read input files. Moreover, the scalability of the pileup algorithm is also poor. Therefore, we designed a new algorithm, named BamPileup, aiming to improve the performance of sequential read, and the new pileup algorithm implemented a parallel read mode based on index. Using this method, each thread can directly read the data start from a specific position. The results of experiments on the Tianhe-2 supercomputer show that, when reading data in a multi-threaded parallel I/O way, the processing time of algorithm is reduced to 3.9 s and the application program can achieve a speedup up to 100×. Moreover, the scalability of the new algorithm is also satisfying.

Computing Gröbner and Involutive Bases for Linear Systems of Difference Equations

NASA Astrophysics Data System (ADS)

Yanovich, Denis

2018-02-01

The computation of involutive bases and Gröbner bases for linear systems of difference equations is solved and its importance for physical and mathematical problems is discussed. The algorithm and issues concerning its implementation in C are presented and calculation times are compared with the competing programs. The paper ends with consideration on the parallel version of this implementation and its scalability.

Trainable hardware for dynamical computing using error backpropagation through physical media.

PubMed

Hermans, Michiel; Burm, Michaël; Van Vaerenbergh, Thomas; Dambre, Joni; Bienstman, Peter

2015-03-24

Neural networks are currently implemented on digital Von Neumann machines, which do not fully leverage their intrinsic parallelism. We demonstrate how to use a novel class of reconfigurable dynamical systems for analogue information processing, mitigating this problem. Our generic hardware platform for dynamic, analogue computing consists of a reciprocal linear dynamical system with nonlinear feedback. Thanks to reciprocity, a ubiquitous property of many physical phenomena like the propagation of light and sound, the error backpropagation-a crucial step for tuning such systems towards a specific task-can happen in hardware. This can potentially speed up the optimization process significantly, offering important benefits for the scalability of neuro-inspired hardware. In this paper, we show, using one experimentally validated and one conceptual example, that such systems may provide a straightforward mechanism for constructing highly scalable, fully dynamical analogue computers.

Trainable hardware for dynamical computing using error backpropagation through physical media

NASA Astrophysics Data System (ADS)

Hermans, Michiel; Burm, Michaël; van Vaerenbergh, Thomas; Dambre, Joni; Bienstman, Peter

2015-03-01

Neural networks are currently implemented on digital Von Neumann machines, which do not fully leverage their intrinsic parallelism. We demonstrate how to use a novel class of reconfigurable dynamical systems for analogue information processing, mitigating this problem. Our generic hardware platform for dynamic, analogue computing consists of a reciprocal linear dynamical system with nonlinear feedback. Thanks to reciprocity, a ubiquitous property of many physical phenomena like the propagation of light and sound, the error backpropagation—a crucial step for tuning such systems towards a specific task—can happen in hardware. This can potentially speed up the optimization process significantly, offering important benefits for the scalability of neuro-inspired hardware. In this paper, we show, using one experimentally validated and one conceptual example, that such systems may provide a straightforward mechanism for constructing highly scalable, fully dynamical analogue computers.

Scalable, High-performance 3D Imaging Software Platform: System Architecture and Application to Virtual Colonoscopy

PubMed Central

Yoshida, Hiroyuki; Wu, Yin; Cai, Wenli; Brett, Bevin

2013-01-01

One of the key challenges in three-dimensional (3D) medical imaging is to enable the fast turn-around time, which is often required for interactive or real-time response. This inevitably requires not only high computational power but also high memory bandwidth due to the massive amount of data that need to be processed. In this work, we have developed a software platform that is designed to support high-performance 3D medical image processing for a wide range of applications using increasingly available and affordable commodity computing systems: multi-core, clusters, and cloud computing systems. To achieve scalable, high-performance computing, our platform (1) employs size-adaptive, distributable block volumes as a core data structure for efficient parallelization of a wide range of 3D image processing algorithms; (2) supports task scheduling for efficient load distribution and balancing; and (3) consists of a layered parallel software libraries that allow a wide range of medical applications to share the same functionalities. We evaluated the performance of our platform by applying it to an electronic cleansing system in virtual colonoscopy, with initial experimental results showing a 10 times performance improvement on an 8-core workstation over the original sequential implementation of the system. PMID:23366803

Performance Models for the Spike Banded Linear System Solver

DOE PAGES

Manguoglu, Murat; Saied, Faisal; Sameh, Ahmed; ...

2011-01-01

With availability of large-scale parallel platforms comprised of tens-of-thousands of processors and beyond, there is significant impetus for the development of scalable parallel sparse linear system solvers and preconditioners. An integral part of this design process is the development of performance models capable of predicting performance and providing accurate cost models for the solvers and preconditioners. There has been some work in the past on characterizing performance of the iterative solvers themselves. In this paper, we investigate the problem of characterizing performance and scalability of banded preconditioners. Recent work has demonstrated the superior convergence properties and robustness of banded preconditioners,more » compared to state-of-the-art ILU family of preconditioners as well as algebraic multigrid preconditioners. Furthermore, when used in conjunction with efficient banded solvers, banded preconditioners are capable of significantly faster time-to-solution. Our banded solver, the Truncated Spike algorithm is specifically designed for parallel performance and tolerance to deep memory hierarchies. Its regular structure is also highly amenable to accurate performance characterization. Using these characteristics, we derive the following results in this paper: (i) we develop parallel formulations of the Truncated Spike solver, (ii) we develop a highly accurate pseudo-analytical parallel performance model for our solver, (iii) we show excellent predication capabilities of our model – based on which we argue the high scalability of our solver. Our pseudo-analytical performance model is based on analytical performance characterization of each phase of our solver. These analytical models are then parameterized using actual runtime information on target platforms. An important consequence of our performance models is that they reveal underlying performance bottlenecks in both serial and parallel formulations. All of our results are validated on diverse heterogeneous multiclusters – platforms for which performance prediction is particularly challenging. Finally, we provide predict the scalability of the Spike algorithm using up to 65,536 cores with our model. In this paper we extend the results presented in the Ninth International Symposium on Parallel and Distributed Computing.« less

Optimized scalable network switch

DOEpatents

Blumrich, Matthias A [Ridgefield, CT; Chen, Dong [Croton On Hudson, NY; Coteus, Paul W [Yorktown Heights, NY; Gara, Alan G [Mount Kisco, NY; Giampapa, Mark E [Irvington, NY; Heidelberger, Philip [Cortlandt Manor, NY; Steinmacher-Burow, Burkhard D [Mount Kisco, NY; Takken, Todd E [Mount Kisco, NY; Vranas, Pavlos M [Bedford Hills, NY

2007-12-04

In a massively parallel computing system having a plurality of nodes configured in m multi-dimensions, each node including a computing device, a method for routing packets towards their destination nodes is provided which includes generating at least one of a 2m plurality of compact bit vectors containing information derived from downstream nodes. A multilevel arbitration process in which downstream information stored in the compact vectors, such as link status information and fullness of downstream buffers, is used to determine a preferred direction and virtual channel for packet transmission. Preferred direction ranges are encoded and virtual channels are selected by examining the plurality of compact bit vectors. This dynamic routing method eliminates the necessity of routing tables, thus enhancing scalability of the switch.

Optimized scalable network switch

DOEpatents

Blumrich, Matthias A.; Chen, Dong; Coteus, Paul W.

2010-02-23

In a massively parallel computing system having a plurality of nodes configured in m multi-dimensions, each node including a computing device, a method for routing packets towards their destination nodes is provided which includes generating at least one of a 2m plurality of compact bit vectors containing information derived from downstream nodes. A multilevel arbitration process in which downstream information stored in the compact vectors, such as link status information and fullness of downstream buffers, is used to determine a preferred direction and virtual channel for packet transmission. Preferred direction ranges are encoded and virtual channels are selected by examining the plurality of compact bit vectors. This dynamic routing method eliminates the necessity of routing tables, thus enhancing scalability of the switch.

Effect of asynchrony on numerical simulations of fluid flow phenomena

NASA Astrophysics Data System (ADS)

Konduri, Aditya; Mahoney, Bryan; Donzis, Diego

2015-11-01

Designing scalable CFD codes on massively parallel computers is a challenge. This is mainly due to the large number of communications between processing elements (PEs) and their synchronization, leading to idling of PEs. Indeed, communication will likely be the bottleneck in the scalability of codes on Exascale machines. Our recent work on asynchronous computing for PDEs based on finite-differences has shown that it is possible to relax synchronization between PEs at a mathematical level. Computations then proceed regardless of the status of communication, reducing the idle time of PEs and improving the scalability. However, accuracy of the schemes is greatly affected. We have proposed asynchrony-tolerant (AT) schemes to address this issue. In this work, we study the effect of asynchrony on the solution of fluid flow problems using standard and AT schemes. We show that asynchrony creates additional scales with low energy content. The specific wavenumbers affected can be shown to be due to two distinct effects: the randomness in the arrival of messages and the corresponding switching between schemes. Understanding these errors allow us to effectively control them, rendering the method's feasibility in solving turbulent flows at realistic conditions on future computing systems.

Using domain decomposition in the multigrid NAS parallel benchmark on the Fujitsu VPP500

DOE Office of Scientific and Technical Information (OSTI.GOV)

Wang, J.C.H.; Lung, H.; Katsumata, Y.

1995-12-01

In this paper, we demonstrate how domain decomposition can be applied to the multigrid algorithm to convert the code for MPP architectures. We also discuss the performance and scalability of this implementation on the new product line of Fujitsu`s vector parallel computer, VPP500. This computer has Fujitsu`s well-known vector processor as the PE each rated at 1.6 C FLOPS. The high speed crossbar network rated at 800 MB/s provides the inter-PE communication. The results show that the physical domain decomposition is the best way to solve MG problems on VPP500.

Numerical characteristics of quantum computer simulation

NASA Astrophysics Data System (ADS)

Chernyavskiy, A.; Khamitov, K.; Teplov, A.; Voevodin, V.; Voevodin, Vl.

2016-12-01

The simulation of quantum circuits is significantly important for the implementation of quantum information technologies. The main difficulty of such modeling is the exponential growth of dimensionality, thus the usage of modern high-performance parallel computations is relevant. As it is well known, arbitrary quantum computation in circuit model can be done by only single- and two-qubit gates, and we analyze the computational structure and properties of the simulation of such gates. We investigate the fact that the unique properties of quantum nature lead to the computational properties of the considered algorithms: the quantum parallelism make the simulation of quantum gates highly parallel, and on the other hand, quantum entanglement leads to the problem of computational locality during simulation. We use the methodology of the AlgoWiki project (algowiki-project.org) to analyze the algorithm. This methodology consists of theoretical (sequential and parallel complexity, macro structure, and visual informational graph) and experimental (locality and memory access, scalability and more specific dynamic characteristics) parts. Experimental part was made by using the petascale Lomonosov supercomputer (Moscow State University, Russia). We show that the simulation of quantum gates is a good base for the research and testing of the development methods for data intense parallel software, and considered methodology of the analysis can be successfully used for the improvement of the algorithms in quantum information science.

Evolution of a minimal parallel programming model

DOE PAGES

Lusk, Ewing; Butler, Ralph; Pieper, Steven C.

2017-04-30

Here, we take a historical approach to our presentation of self-scheduled task parallelism, a programming model with its origins in early irregular and nondeterministic computations encountered in automated theorem proving and logic programming. We show how an extremely simple task model has evolved into a system, asynchronous dynamic load balancing (ADLB), and a scalable implementation capable of supporting sophisticated applications on today’s (and tomorrow’s) largest supercomputers; and we illustrate the use of ADLB with a Green’s function Monte Carlo application, a modern, mature nuclear physics code in production use. Our lesson is that by surrendering a certain amount of generalitymore » and thus applicability, a minimal programming model (in terms of its basic concepts and the size of its application programmer interface) can achieve extreme scalability without introducing complexity.« less

Scalable software architecture for on-line multi-camera video processing

NASA Astrophysics Data System (ADS)

Camplani, Massimo; Salgado, Luis

2011-03-01

In this paper we present a scalable software architecture for on-line multi-camera video processing, that guarantees a good trade off between computational power, scalability and flexibility. The software system is modular and its main blocks are the Processing Units (PUs), and the Central Unit. The Central Unit works as a supervisor of the running PUs and each PU manages the acquisition phase and the processing phase. Furthermore, an approach to easily parallelize the desired processing application has been presented. In this paper, as case study, we apply the proposed software architecture to a multi-camera system in order to efficiently manage multiple 2D object detection modules in a real-time scenario. System performance has been evaluated under different load conditions such as number of cameras and image sizes. The results show that the software architecture scales well with the number of camera and can easily works with different image formats respecting the real time constraints. Moreover, the parallelization approach can be used in order to speed up the processing tasks with a low level of overhead.

High-Performance Computation of Distributed-Memory Parallel 3D Voronoi and Delaunay Tessellation

DOE Office of Scientific and Technical Information (OSTI.GOV)

Peterka, Tom; Morozov, Dmitriy; Phillips, Carolyn

2014-11-14

Computing a Voronoi or Delaunay tessellation from a set of points is a core part of the analysis of many simulated and measured datasets: N-body simulations, molecular dynamics codes, and LIDAR point clouds are just a few examples. Such computational geometry methods are common in data analysis and visualization; but as the scale of simulations and observations surpasses billions of particles, the existing serial and shared-memory algorithms no longer suffice. A distributed-memory scalable parallel algorithm is the only feasible approach. The primary contribution of this paper is a new parallel Delaunay and Voronoi tessellation algorithm that automatically determines which neighbormore » points need to be exchanged among the subdomains of a spatial decomposition. Other contributions include periodic and wall boundary conditions, comparison of our method using two popular serial libraries, and application to numerous science datasets.« less

A Linked-Cell Domain Decomposition Method for Molecular Dynamics Simulation on a Scalable Multiprocessor

DOE PAGES

Yang, L. H.; Brooks III, E. D.; Belak, J.

1992-01-01

A molecular dynamics algorithm for performing large-scale simulations using the Parallel C Preprocessor (PCP) programming paradigm on the BBN TC2000, a massively parallel computer, is discussed. The algorithm uses a linked-cell data structure to obtain the near neighbors of each atom as time evoles. Each processor is assigned to a geometric domain containing many subcells and the storage for that domain is private to the processor. Within this scheme, the interdomain (i.e., interprocessor) communication is minimized.

Towards a large-scale scalable adaptive heart model using shallow tree meshes

NASA Astrophysics Data System (ADS)

Krause, Dorian; Dickopf, Thomas; Potse, Mark; Krause, Rolf

2015-10-01

Electrophysiological heart models are sophisticated computational tools that place high demands on the computing hardware due to the high spatial resolution required to capture the steep depolarization front. To address this challenge, we present a novel adaptive scheme for resolving the deporalization front accurately using adaptivity in space. Our adaptive scheme is based on locally structured meshes. These tensor meshes in space are organized in a parallel forest of trees, which allows us to resolve complicated geometries and to realize high variations in the local mesh sizes with a minimal memory footprint in the adaptive scheme. We discuss both a non-conforming mortar element approximation and a conforming finite element space and present an efficient technique for the assembly of the respective stiffness matrices using matrix representations of the inclusion operators into the product space on the so-called shallow tree meshes. We analyzed the parallel performance and scalability for a two-dimensional ventricle slice as well as for a full large-scale heart model. Our results demonstrate that the method has good performance and high accuracy.

Multi-jagged: A scalable parallel spatial partitioning algorithm

DOE PAGES

Deveci, Mehmet; Rajamanickam, Sivasankaran; Devine, Karen D.; ...

2015-03-18

Geometric partitioning is fast and effective for load-balancing dynamic applications, particularly those requiring geometric locality of data (particle methods, crash simulations). We present, to our knowledge, the first parallel implementation of a multidimensional-jagged geometric partitioner. In contrast to the traditional recursive coordinate bisection algorithm (RCB), which recursively bisects subdomains perpendicular to their longest dimension until the desired number of parts is obtained, our algorithm does recursive multi-section with a given number of parts in each dimension. By computing multiple cut lines concurrently and intelligently deciding when to migrate data while computing the partition, we minimize data movement compared to efficientmore » implementations of recursive bisection. We demonstrate the algorithm's scalability and quality relative to the RCB implementation in Zoltan on both real and synthetic datasets. Our experiments show that the proposed algorithm performs and scales better than RCB in terms of run-time without degrading the load balance. Lastly, our implementation partitions 24 billion points into 65,536 parts within a few seconds and exhibits near perfect weak scaling up to 6K cores.« less

Scalable real space pseudopotential density functional codes for materials in the exascale regime

NASA Astrophysics Data System (ADS)

Lena, Charles; Chelikowsky, James; Schofield, Grady; Biller, Ariel; Kronik, Leeor; Saad, Yousef; Deslippe, Jack

Real-space pseudopotential density functional theory has proven to be an efficient method for computing the properties of matter in many different states and geometries, including liquids, wires, slabs, and clusters with and without spin polarization. Fully self-consistent solutions using this approach have been routinely obtained for systems with thousands of atoms. Yet, there are many systems of notable larger sizes where quantum mechanical accuracy is desired, but scalability proves to be a hindrance. Such systems include large biological molecules, complex nanostructures, or mismatched interfaces. We will present an overview of our new massively parallel algorithms, which offer improved scalability in preparation for exascale supercomputing. We will illustrate these algorithms by considering the electronic structure of a Si nanocrystal exceeding 104 atoms. Support provided by the SciDAC program, Department of Energy, Office of Science, Advanced Scientific Computing Research and Basic Energy Sciences. Grant Numbers DE-SC0008877 (Austin) and DE-FG02-12ER4 (Berkeley).

Parallel volume ray-casting for unstructured-grid data on distributed-memory architectures

NASA Technical Reports Server (NTRS)

Ma, Kwan-Liu

1995-01-01

As computing technology continues to advance, computational modeling of scientific and engineering problems produces data of increasing complexity: large in size and unstructured in shape. Volume visualization of such data is a challenging problem. This paper proposes a distributed parallel solution that makes ray-casting volume rendering of unstructured-grid data practical. Both the data and the rendering process are distributed among processors. At each processor, ray-casting of local data is performed independent of the other processors. The global image composing processes, which require inter-processor communication, are overlapped with the local ray-casting processes to achieve maximum parallel efficiency. This algorithm differs from previous ones in four ways: it is completely distributed, less view-dependent, reasonably scalable, and flexible. Without using dynamic load balancing, test results on the Intel Paragon using from two to 128 processors show, on average, about 60% parallel efficiency.

Three-Dimensional High-Lift Analysis Using a Parallel Unstructured Multigrid Solver

NASA Technical Reports Server (NTRS)

Mavriplis, Dimitri J.

1998-01-01

A directional implicit unstructured agglomeration multigrid solver is ported to shared and distributed memory massively parallel machines using the explicit domain-decomposition and message-passing approach. Because the algorithm operates on local implicit lines in the unstructured mesh, special care is required in partitioning the problem for parallel computing. A weighted partitioning strategy is described which avoids breaking the implicit lines across processor boundaries, while incurring minimal additional communication overhead. Good scalability is demonstrated on a 128 processor SGI Origin 2000 machine and on a 512 processor CRAY T3E machine for reasonably fine grids. The feasibility of performing large-scale unstructured grid calculations with the parallel multigrid algorithm is demonstrated by computing the flow over a partial-span flap wing high-lift geometry on a highly resolved grid of 13.5 million points in approximately 4 hours of wall clock time on the CRAY T3E.

Parallel computing of a digital hologram and particle searching for microdigital-holographic particle-tracking velocimetry

DOE Office of Scientific and Technical Information (OSTI.GOV)

Satake, Shin-ichi; Kanamori, Hiroyuki; Kunugi, Tomoaki

2007-02-01

We have developed a parallel algorithm for microdigital-holographic particle-tracking velocimetry. The algorithm is used in (1) numerical reconstruction of a particle image computer using a digital hologram, and (2) searching for particles. The numerical reconstruction from the digital hologram makes use of the Fresnel diffraction equation and the FFT (fast Fourier transform),whereas the particle search algorithm looks for local maximum graduation in a reconstruction field represented by a 3D matrix. To achieve high performance computing for both calculations (reconstruction and particle search), two memory partitions are allocated to the 3D matrix. In this matrix, the reconstruction part consists of horizontallymore » placed 2D memory partitions on the x-y plane for the FFT, whereas, the particle search part consists of vertically placed 2D memory partitions set along the z axes.Consequently, the scalability can be obtained for the proportion of processor elements,where the benchmarks are carried out for parallel computation by a SGI Altix machine.« less

Parallel Navier-Stokes computations on shared and distributed memory architectures

NASA Technical Reports Server (NTRS)

Hayder, M. Ehtesham; Jayasimha, D. N.; Pillay, Sasi Kumar

1995-01-01

We study a high order finite difference scheme to solve the time accurate flow field of a jet using the compressible Navier-Stokes equations. As part of our ongoing efforts, we have implemented our numerical model on three parallel computing platforms to study the computational, communication, and scalability characteristics. The platforms chosen for this study are a cluster of workstations connected through fast networks (the LACE experimental testbed at NASA Lewis), a shared memory multiprocessor (the Cray YMP), and a distributed memory multiprocessor (the IBM SPI). Our focus in this study is on the LACE testbed. We present some results for the Cray YMP and the IBM SP1 mainly for comparison purposes. On the LACE testbed, we study: (1) the communication characteristics of Ethernet, FDDI, and the ALLNODE networks and (2) the overheads induced by the PVM message passing library used for parallelizing the application. We demonstrate that clustering of workstations is effective and has the potential to be computationally competitive with supercomputers at a fraction of the cost.

Feasibility of optically interconnected parallel processors using wavelength division multiplexing

DOE Office of Scientific and Technical Information (OSTI.GOV)

Deri, R.J.; De Groot, A.J.; Haigh, R.E.

1996-03-01

New national security demands require enhanced computing systems for nearly ab initio simulations of extremely complex systems and analyzing unprecedented quantities of remote sensing data. This computational performance is being sought using parallel processing systems, in which many less powerful processors are ganged together to achieve high aggregate performance. Such systems require increased capability to communicate information between individual processor and memory elements. As it is likely that the limited performance of today`s electronic interconnects will prevent the system from achieving its ultimate performance, there is great interest in using fiber optic technology to improve interconnect communication. However, little informationmore » is available to quantify the requirements on fiber optical hardware technology for this application. Furthermore, we have sought to explore interconnect architectures that use the complete communication richness of the optical domain rather than using optics as a simple replacement for electronic interconnects. These considerations have led us to study the performance of a moderate size parallel processor with optical interconnects using multiple optical wavelengths. We quantify the bandwidth, latency, and concurrency requirements which allow a bus-type interconnect to achieve scalable computing performance using up to 256 nodes, each operating at GFLOP performance. Our key conclusion is that scalable performance, to {approx}150 GFLOPS, is achievable for several scientific codes using an optical bus with a small number of WDM channels (8 to 32), only one WDM channel received per node, and achievable optoelectronic bandwidth and latency requirements. 21 refs. , 10 figs.« less

The Automatic Parallelisation of Scientific Application Codes Using a Computer Aided Parallelisation Toolkit

NASA Technical Reports Server (NTRS)

Ierotheou, C.; Johnson, S.; Leggett, P.; Cross, M.; Evans, E.; Jin, Hao-Qiang; Frumkin, M.; Yan, J.; Biegel, Bryan (Technical Monitor)

2001-01-01

The shared-memory programming model is a very effective way to achieve parallelism on shared memory parallel computers. Historically, the lack of a programming standard for using directives and the rather limited performance due to scalability have affected the take-up of this programming model approach. Significant progress has been made in hardware and software technologies, as a result the performance of parallel programs with compiler directives has also made improvements. The introduction of an industrial standard for shared-memory programming with directives, OpenMP, has also addressed the issue of portability. In this study, we have extended the computer aided parallelization toolkit (developed at the University of Greenwich), to automatically generate OpenMP based parallel programs with nominal user assistance. We outline the way in which loop types are categorized and how efficient OpenMP directives can be defined and placed using the in-depth interprocedural analysis that is carried out by the toolkit. We also discuss the application of the toolkit on the NAS Parallel Benchmarks and a number of real-world application codes. This work not only demonstrates the great potential of using the toolkit to quickly parallelize serial programs but also the good performance achievable on up to 300 processors for hybrid message passing and directive-based parallelizations.

Sublattice parallel replica dynamics.

PubMed

Martínez, Enrique; Uberuaga, Blas P; Voter, Arthur F

2014-06-01

Exascale computing presents a challenge for the scientific community as new algorithms must be developed to take full advantage of the new computing paradigm. Atomistic simulation methods that offer full fidelity to the underlying potential, i.e., molecular dynamics (MD) and parallel replica dynamics, fail to use the whole machine speedup, leaving a region in time and sample size space that is unattainable with current algorithms. In this paper, we present an extension of the parallel replica dynamics algorithm [A. F. Voter, Phys. Rev. B 57, R13985 (1998)] by combining it with the synchronous sublattice approach of Shim and Amar [ and , Phys. Rev. B 71, 125432 (2005)], thereby exploiting event locality to improve the algorithm scalability. This algorithm is based on a domain decomposition in which events happen independently in different regions in the sample. We develop an analytical expression for the speedup given by this sublattice parallel replica dynamics algorithm and compare it with parallel MD and traditional parallel replica dynamics. We demonstrate how this algorithm, which introduces a slight additional approximation of event locality, enables the study of physical systems unreachable with traditional methodologies and promises to better utilize the resources of current high performance and future exascale computers.

Dynamic Load Balancing for Adaptive Computations on Distributed-Memory Machines

NASA Technical Reports Server (NTRS)

1999-01-01

Dynamic load balancing is central to adaptive mesh-based computations on large-scale parallel computers. The principal investigator has investigated various issues on the dynamic load balancing problem under NASA JOVE and JAG rants. The major accomplishments of the project are two graph partitioning algorithms and a load balancing framework. The S-HARP dynamic graph partitioner is known to be the fastest among the known dynamic graph partitioners to date. It can partition a graph of over 100,000 vertices in 0.25 seconds on a 64- processor Cray T3E distributed-memory multiprocessor while maintaining the scalability of over 16-fold speedup. Other known and widely used dynamic graph partitioners take over a second or two while giving low scalability of a few fold speedup on 64 processors. These results have been published in journals and peer-reviewed flagship conferences.

High-performance computational fluid dynamics: a custom-code approach

NASA Astrophysics Data System (ADS)

Fannon, James; Loiseau, Jean-Christophe; Valluri, Prashant; Bethune, Iain; Náraigh, Lennon Ó.

2016-07-01

We introduce a modified and simplified version of the pre-existing fully parallelized three-dimensional Navier-Stokes flow solver known as TPLS. We demonstrate how the simplified version can be used as a pedagogical tool for the study of computational fluid dynamics (CFDs) and parallel computing. TPLS is at its heart a two-phase flow solver, and uses calls to a range of external libraries to accelerate its performance. However, in the present context we narrow the focus of the study to basic hydrodynamics and parallel computing techniques, and the code is therefore simplified and modified to simulate pressure-driven single-phase flow in a channel, using only relatively simple Fortran 90 code with MPI parallelization, but no calls to any other external libraries. The modified code is analysed in order to both validate its accuracy and investigate its scalability up to 1000 CPU cores. Simulations are performed for several benchmark cases in pressure-driven channel flow, including a turbulent simulation, wherein the turbulence is incorporated via the large-eddy simulation technique. The work may be of use to advanced undergraduate and graduate students as an introductory study in CFDs, while also providing insight for those interested in more general aspects of high-performance computing.

Monte Carlo MP2 on Many Graphical Processing Units.

PubMed

Doran, Alexander E; Hirata, So

2016-10-11

In the Monte Carlo second-order many-body perturbation (MC-MP2) method, the long sum-of-product matrix expression of the MP2 energy, whose literal evaluation may be poorly scalable, is recast into a single high-dimensional integral of functions of electron pair coordinates, which is evaluated by the scalable method of Monte Carlo integration. The sampling efficiency is further accelerated by the redundant-walker algorithm, which allows a maximal reuse of electron pairs. Here, a multitude of graphical processing units (GPUs) offers a uniquely ideal platform to expose multilevel parallelism: fine-grain data-parallelism for the redundant-walker algorithm in which millions of threads compute and share orbital amplitudes on each GPU; coarse-grain instruction-parallelism for near-independent Monte Carlo integrations on many GPUs with few and infrequent interprocessor communications. While the efficiency boost by the redundant-walker algorithm on central processing units (CPUs) grows linearly with the number of electron pairs and tends to saturate when the latter exceeds the number of orbitals, on a GPU it grows quadratically before it increases linearly and then eventually saturates at a much larger number of pairs. This is because the orbital constructions are nearly perfectly parallelized on a GPU and thus completed in a near-constant time regardless of the number of pairs. In consequence, an MC-MP2/cc-pVDZ calculation of a benzene dimer is 2700 times faster on 256 GPUs (using 2048 electron pairs) than on two CPUs, each with 8 cores (which can use only up to 256 pairs effectively). We also numerically determine that the cost to achieve a given relative statistical uncertainty in an MC-MP2 energy increases as O(n 3 ) or better with system size n, which may be compared with the O(n 5 ) scaling of the conventional implementation of deterministic MP2. We thus establish the scalability of MC-MP2 with both system and computer sizes.

Scalable Evaluation of Polarization Energy and Associated Forces in Polarizable Molecular Dynamics: II.Towards Massively Parallel Computations using Smooth Particle Mesh Ewald.

PubMed

Lagardère, Louis; Lipparini, Filippo; Polack, Étienne; Stamm, Benjamin; Cancès, Éric; Schnieders, Michael; Ren, Pengyu; Maday, Yvon; Piquemal, Jean-Philip

2014-02-28

In this paper, we present a scalable and efficient implementation of point dipole-based polarizable force fields for molecular dynamics (MD) simulations with periodic boundary conditions (PBC). The Smooth Particle-Mesh Ewald technique is combined with two optimal iterative strategies, namely, a preconditioned conjugate gradient solver and a Jacobi solver in conjunction with the Direct Inversion in the Iterative Subspace for convergence acceleration, to solve the polarization equations. We show that both solvers exhibit very good parallel performances and overall very competitive timings in an energy-force computation needed to perform a MD step. Various tests on large systems are provided in the context of the polarizable AMOEBA force field as implemented in the newly developed Tinker-HP package which is the first implementation for a polarizable model making large scale experiments for massively parallel PBC point dipole models possible. We show that using a large number of cores offers a significant acceleration of the overall process involving the iterative methods within the context of spme and a noticeable improvement of the memory management giving access to very large systems (hundreds of thousands of atoms) as the algorithm naturally distributes the data on different cores. Coupled with advanced MD techniques, gains ranging from 2 to 3 orders of magnitude in time are now possible compared to non-optimized, sequential implementations giving new directions for polarizable molecular dynamics in periodic boundary conditions using massively parallel implementations.

Scalable Evaluation of Polarization Energy and Associated Forces in Polarizable Molecular Dynamics: II.Towards Massively Parallel Computations using Smooth Particle Mesh Ewald

PubMed Central

Lagardère, Louis; Lipparini, Filippo; Polack, Étienne; Stamm, Benjamin; Cancès, Éric; Schnieders, Michael; Ren, Pengyu; Maday, Yvon; Piquemal, Jean-Philip

2015-01-01

In this paper, we present a scalable and efficient implementation of point dipole-based polarizable force fields for molecular dynamics (MD) simulations with periodic boundary conditions (PBC). The Smooth Particle-Mesh Ewald technique is combined with two optimal iterative strategies, namely, a preconditioned conjugate gradient solver and a Jacobi solver in conjunction with the Direct Inversion in the Iterative Subspace for convergence acceleration, to solve the polarization equations. We show that both solvers exhibit very good parallel performances and overall very competitive timings in an energy-force computation needed to perform a MD step. Various tests on large systems are provided in the context of the polarizable AMOEBA force field as implemented in the newly developed Tinker-HP package which is the first implementation for a polarizable model making large scale experiments for massively parallel PBC point dipole models possible. We show that using a large number of cores offers a significant acceleration of the overall process involving the iterative methods within the context of spme and a noticeable improvement of the memory management giving access to very large systems (hundreds of thousands of atoms) as the algorithm naturally distributes the data on different cores. Coupled with advanced MD techniques, gains ranging from 2 to 3 orders of magnitude in time are now possible compared to non-optimized, sequential implementations giving new directions for polarizable molecular dynamics in periodic boundary conditions using massively parallel implementations. PMID:26512230

Parallel algorithms for large-scale biological sequence alignment on Xeon-Phi based clusters.

PubMed

Lan, Haidong; Chan, Yuandong; Xu, Kai; Schmidt, Bertil; Peng, Shaoliang; Liu, Weiguo

2016-07-19

Computing alignments between two or more sequences are common operations frequently performed in computational molecular biology. The continuing growth of biological sequence databases establishes the need for their efficient parallel implementation on modern accelerators. This paper presents new approaches to high performance biological sequence database scanning with the Smith-Waterman algorithm and the first stage of progressive multiple sequence alignment based on the ClustalW heuristic on a Xeon Phi-based compute cluster. Our approach uses a three-level parallelization scheme to take full advantage of the compute power available on this type of architecture; i.e. cluster-level data parallelism, thread-level coarse-grained parallelism, and vector-level fine-grained parallelism. Furthermore, we re-organize the sequence datasets and use Xeon Phi shuffle operations to improve I/O efficiency. Evaluations show that our method achieves a peak overall performance up to 220 GCUPS for scanning real protein sequence databanks on a single node consisting of two Intel E5-2620 CPUs and two Intel Xeon Phi 7110P cards. It also exhibits good scalability in terms of sequence length and size, and number of compute nodes for both database scanning and multiple sequence alignment. Furthermore, the achieved performance is highly competitive in comparison to optimized Xeon Phi and GPU implementations. Our implementation is available at https://github.com/turbo0628/LSDBS-mpi .

A Robust and Scalable Software Library for Parallel Adaptive Refinement on Unstructured Meshes

NASA Technical Reports Server (NTRS)

Lou, John Z.; Norton, Charles D.; Cwik, Thomas A.

1999-01-01

The design and implementation of Pyramid, a software library for performing parallel adaptive mesh refinement (PAMR) on unstructured meshes, is described. This software library can be easily used in a variety of unstructured parallel computational applications, including parallel finite element, parallel finite volume, and parallel visualization applications using triangular or tetrahedral meshes. The library contains a suite of well-designed and efficiently implemented modules that perform operations in a typical PAMR process. Among these are mesh quality control during successive parallel adaptive refinement (typically guided by a local-error estimator), parallel load-balancing, and parallel mesh partitioning using the ParMeTiS partitioner. The Pyramid library is implemented in Fortran 90 with an interface to the Message-Passing Interface (MPI) library, supporting code efficiency, modularity, and portability. An EM waveguide filter application, adaptively refined using the Pyramid library, is illustrated.

Fast and Scalable Computation of the Forward and Inverse Discrete Periodic Radon Transform.

PubMed

Carranza, Cesar; Llamocca, Daniel; Pattichis, Marios

2016-01-01

The discrete periodic radon transform (DPRT) has extensively been used in applications that involve image reconstructions from projections. Beyond classic applications, the DPRT can also be used to compute fast convolutions that avoids the use of floating-point arithmetic associated with the use of the fast Fourier transform. Unfortunately, the use of the DPRT has been limited by the need to compute a large number of additions and the need for a large number of memory accesses. This paper introduces a fast and scalable approach for computing the forward and inverse DPRT that is based on the use of: a parallel array of fixed-point adder trees; circular shift registers to remove the need for accessing external memory components when selecting the input data for the adder trees; an image block-based approach to DPRT computation that can fit the proposed architecture to available resources; and fast transpositions that are computed in one or a few clock cycles that do not depend on the size of the input image. As a result, for an N × N image (N prime), the proposed approach can compute up to N(2) additions per clock cycle. Compared with the previous approaches, the scalable approach provides the fastest known implementations for different amounts of computational resources. For example, for a 251×251 image, for approximately 25% fewer flip-flops than required for a systolic implementation, we have that the scalable DPRT is computed 36 times faster. For the fastest case, we introduce optimized just 2N + ⌈log(2) N⌉ + 1 and 2N + 3 ⌈log(2) N⌉ + B + 2 cycles, architectures that can compute the DPRT and its inverse in respectively, where B is the number of bits used to represent each input pixel. On the other hand, the scalable DPRT approach requires more 1-b additions than for the systolic implementation and provides a tradeoff between speed and additional 1-b additions. All of the proposed DPRT architectures were implemented in VHSIC Hardware Description Language (VHDL) and validated using an Field-Programmable Gate Array (FPGA) implementation.

Discrete Event Modeling and Massively Parallel Execution of Epidemic Outbreak Phenomena

DOE Office of Scientific and Technical Information (OSTI.GOV)

Perumalla, Kalyan S; Seal, Sudip K

2011-01-01

In complex phenomena such as epidemiological outbreaks, the intensity of inherent feedback effects and the significant role of transients in the dynamics make simulation the only effective method for proactive, reactive or post-facto analysis. The spatial scale, runtime speed, and behavioral detail needed in detailed simulations of epidemic outbreaks make it necessary to use large-scale parallel processing. Here, an optimistic parallel execution of a new discrete event formulation of a reaction-diffusion simulation model of epidemic propagation is presented to facilitate in dramatically increasing the fidelity and speed by which epidemiological simulations can be performed. Rollback support needed during optimistic parallelmore » execution is achieved by combining reverse computation with a small amount of incremental state saving. Parallel speedup of over 5,500 and other runtime performance metrics of the system are observed with weak-scaling execution on a small (8,192-core) Blue Gene / P system, while scalability with a weak-scaling speedup of over 10,000 is demonstrated on 65,536 cores of a large Cray XT5 system. Scenarios representing large population sizes exceeding several hundreds of millions of individuals in the largest cases are successfully exercised to verify model scalability.« less

The TeraShake Computational Platform for Large-Scale Earthquake Simulations

NASA Astrophysics Data System (ADS)

Cui, Yifeng; Olsen, Kim; Chourasia, Amit; Moore, Reagan; Maechling, Philip; Jordan, Thomas

Geoscientific and computer science researchers with the Southern California Earthquake Center (SCEC) are conducting a large-scale, physics-based, computationally demanding earthquake system science research program with the goal of developing predictive models of earthquake processes. The computational demands of this program continue to increase rapidly as these researchers seek to perform physics-based numerical simulations of earthquake processes for larger meet the needs of this research program, a multiple-institution team coordinated by SCEC has integrated several scientific codes into a numerical modeling-based research tool we call the TeraShake computational platform (TSCP). A central component in the TSCP is a highly scalable earthquake wave propagation simulation program called the TeraShake anelastic wave propagation (TS-AWP) code. In this chapter, we describe how we extended an existing, stand-alone, wellvalidated, finite-difference, anelastic wave propagation modeling code into the highly scalable and widely used TS-AWP and then integrated this code into the TeraShake computational platform that provides end-to-end (initialization to analysis) research capabilities. We also describe the techniques used to enhance the TS-AWP parallel performance on TeraGrid supercomputers, as well as the TeraShake simulations phases including input preparation, run time, data archive management, and visualization. As a result of our efforts to improve its parallel efficiency, the TS-AWP has now shown highly efficient strong scaling on over 40K processors on IBM’s BlueGene/L Watson computer. In addition, the TSCP has developed into a computational system that is useful to many members of the SCEC community for performing large-scale earthquake simulations.

Large-scale parallel lattice Boltzmann-cellular automaton model of two-dimensional dendritic growth

NASA Astrophysics Data System (ADS)

Jelinek, Bohumir; Eshraghi, Mohsen; Felicelli, Sergio; Peters, John F.

2014-03-01

An extremely scalable lattice Boltzmann (LB)-cellular automaton (CA) model for simulations of two-dimensional (2D) dendritic solidification under forced convection is presented. The model incorporates effects of phase change, solute diffusion, melt convection, and heat transport. The LB model represents the diffusion, convection, and heat transfer phenomena. The dendrite growth is driven by a difference between actual and equilibrium liquid composition at the solid-liquid interface. The CA technique is deployed to track the new interface cells. The computer program was parallelized using the Message Passing Interface (MPI) technique. Parallel scaling of the algorithm was studied and major scalability bottlenecks were identified. Efficiency loss attributable to the high memory bandwidth requirement of the algorithm was observed when using multiple cores per processor. Parallel writing of the output variables of interest was implemented in the binary Hierarchical Data Format 5 (HDF5) to improve the output performance, and to simplify visualization. Calculations were carried out in single precision arithmetic without significant loss in accuracy, resulting in 50% reduction of memory and computational time requirements. The presented solidification model shows a very good scalability up to centimeter size domains, including more than ten million of dendrites. Catalogue identifier: AEQZ_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEQZ_v1_0.html Program obtainable from: CPC Program Library, Queen’s University, Belfast, UK Licensing provisions: Standard CPC license, http://cpc.cs.qub.ac.uk/licence/licence.html No. of lines in distributed program, including test data, etc.: 29,767 No. of bytes in distributed program, including test data, etc.: 3131,367 Distribution format: tar.gz Programming language: Fortran 90. Computer: Linux PC and clusters. Operating system: Linux. Has the code been vectorized or parallelized?: Yes. Program is parallelized using MPI. Number of processors used: 1-50,000 RAM: Memory requirements depend on the grid size Classification: 6.5, 7.7. External routines: MPI (http://www.mcs.anl.gov/research/projects/mpi/), HDF5 (http://www.hdfgroup.org/HDF5/) Nature of problem: Dendritic growth in undercooled Al-3 wt% Cu alloy melt under forced convection. Solution method: The lattice Boltzmann model solves the diffusion, convection, and heat transfer phenomena. The cellular automaton technique is deployed to track the solid/liquid interface. Restrictions: Heat transfer is calculated uncoupled from the fluid flow. Thermal diffusivity is constant. Unusual features: Novel technique, utilizing periodic duplication of a pre-grown “incubation” domain, is applied for the scaleup test. Running time: Running time varies from minutes to days depending on the domain size and number of computational cores.

A Metascalable Computing Framework for Large Spatiotemporal-Scale Atomistic Simulations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Nomura, K; Seymour, R; Wang, W

2009-02-17

A metascalable (or 'design once, scale on new architectures') parallel computing framework has been developed for large spatiotemporal-scale atomistic simulations of materials based on spatiotemporal data locality principles, which is expected to scale on emerging multipetaflops architectures. The framework consists of: (1) an embedded divide-and-conquer (EDC) algorithmic framework based on spatial locality to design linear-scaling algorithms for high complexity problems; (2) a space-time-ensemble parallel (STEP) approach based on temporal locality to predict long-time dynamics, while introducing multiple parallelization axes; and (3) a tunable hierarchical cellular decomposition (HCD) parallelization framework to map these O(N) algorithms onto a multicore cluster based onmore » hybrid implementation combining message passing and critical section-free multithreading. The EDC-STEP-HCD framework exposes maximal concurrency and data locality, thereby achieving: (1) inter-node parallel efficiency well over 0.95 for 218 billion-atom molecular-dynamics and 1.68 trillion electronic-degrees-of-freedom quantum-mechanical simulations on 212,992 IBM BlueGene/L processors (superscalability); (2) high intra-node, multithreading parallel efficiency (nanoscalability); and (3) nearly perfect time/ensemble parallel efficiency (eon-scalability). The spatiotemporal scale covered by MD simulation on a sustained petaflops computer per day (i.e. petaflops {center_dot} day of computing) is estimated as NT = 2.14 (e.g. N = 2.14 million atoms for T = 1 microseconds).« less

DOE Office of Scientific and Technical Information (OSTI.GOV)

Jain, Atul K.

The overall objectives of this DOE funded project is to combine scientific and computational challenges in climate modeling by expanding our understanding of the biogeophysical-biogeochemical processes and their interactions in the northern high latitudes (NHLs) using an earth system modeling (ESM) approach, and by adopting an adaptive parallel runtime system in an ESM to achieve efficient and scalable climate simulations through improved load balancing algorithms.

Applications Development for a Parallel COTS Spaceborne Computer

NASA Technical Reports Server (NTRS)

Katz, Daniel S.; Springer, Paul L.; Granat, Robert; Turmon, Michael

2000-01-01

This presentation reviews the Remote Exploration and Experimentation Project (REE) program for utilization of scalable supercomputing technology in space. The implementation of REE will be the use of COTS hardware and software to the maximum extent possible, keeping overhead low. Since COTS systems will be used, with little or no special modification, there will be significant cost reduction.

Cross-scale efficient tensor contractions for coupled cluster computations through multiple programming model backends

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ibrahim, Khaled Z.; Epifanovsky, Evgeny; Williams, Samuel

Coupled-cluster methods provide highly accurate models of molecular structure through explicit numerical calculation of tensors representing the correlation between electrons. These calculations are dominated by a sequence of tensor contractions, motivating the development of numerical libraries for such operations. While based on matrix–matrix multiplication, these libraries are specialized to exploit symmetries in the molecular structure and in electronic interactions, and thus reduce the size of the tensor representation and the complexity of contractions. The resulting algorithms are irregular and their parallelization has been previously achieved via the use of dynamic scheduling or specialized data decompositions. We introduce our efforts tomore » extend the Libtensor framework to work in the distributed memory environment in a scalable and energy-efficient manner. We achieve up to 240× speedup compared with the optimized shared memory implementation of Libtensor. We attain scalability to hundreds of thousands of compute cores on three distributed-memory architectures (Cray XC30 and XC40, and IBM Blue Gene/Q), and on a heterogeneous GPU-CPU system (Cray XK7). As the bottlenecks shift from being compute-bound DGEMM's to communication-bound collectives as the size of the molecular system scales, we adopt two radically different parallelization approaches for handling load-imbalance, tasking and bulk synchronous models. Nevertheless, we preserve a unified interface to both programming models to maintain the productivity of computational quantum chemists.« less

Cross-scale efficient tensor contractions for coupled cluster computations through multiple programming model backends

DOE PAGES

Ibrahim, Khaled Z.; Epifanovsky, Evgeny; Williams, Samuel; ...

2017-03-08

Coupled-cluster methods provide highly accurate models of molecular structure through explicit numerical calculation of tensors representing the correlation between electrons. These calculations are dominated by a sequence of tensor contractions, motivating the development of numerical libraries for such operations. While based on matrix–matrix multiplication, these libraries are specialized to exploit symmetries in the molecular structure and in electronic interactions, and thus reduce the size of the tensor representation and the complexity of contractions. The resulting algorithms are irregular and their parallelization has been previously achieved via the use of dynamic scheduling or specialized data decompositions. We introduce our efforts tomore » extend the Libtensor framework to work in the distributed memory environment in a scalable and energy-efficient manner. We achieve up to 240× speedup compared with the optimized shared memory implementation of Libtensor. We attain scalability to hundreds of thousands of compute cores on three distributed-memory architectures (Cray XC30 and XC40, and IBM Blue Gene/Q), and on a heterogeneous GPU-CPU system (Cray XK7). As the bottlenecks shift from being compute-bound DGEMM's to communication-bound collectives as the size of the molecular system scales, we adopt two radically different parallelization approaches for handling load-imbalance, tasking and bulk synchronous models. Nevertheless, we preserve a unified interface to both programming models to maintain the productivity of computational quantum chemists.« less

GLAD: a system for developing and deploying large-scale bioinformatics grid.

PubMed

Teo, Yong-Meng; Wang, Xianbing; Ng, Yew-Kwong

2005-03-01

Grid computing is used to solve large-scale bioinformatics problems with gigabytes database by distributing the computation across multiple platforms. Until now in developing bioinformatics grid applications, it is extremely tedious to design and implement the component algorithms and parallelization techniques for different classes of problems, and to access remotely located sequence database files of varying formats across the grid. In this study, we propose a grid programming toolkit, GLAD (Grid Life sciences Applications Developer), which facilitates the development and deployment of bioinformatics applications on a grid. GLAD has been developed using ALiCE (Adaptive scaLable Internet-based Computing Engine), a Java-based grid middleware, which exploits the task-based parallelism. Two bioinformatics benchmark applications, such as distributed sequence comparison and distributed progressive multiple sequence alignment, have been developed using GLAD.

CFD Research, Parallel Computation and Aerodynamic Optimization

NASA Technical Reports Server (NTRS)

Ryan, James S.

1995-01-01

During the last five years, CFD has matured substantially. Pure CFD research remains to be done, but much of the focus has shifted to integration of CFD into the design process. The work under these cooperative agreements reflects this trend. The recent work, and work which is planned, is designed to enhance the competitiveness of the US aerospace industry. CFD and optimization approaches are being developed and tested, so that the industry can better choose which methods to adopt in their design processes. The range of computer architectures has been dramatically broadened, as the assumption that only huge vector supercomputers could be useful has faded. Today, researchers and industry can trade off time, cost, and availability, choosing vector supercomputers, scalable parallel architectures, networked workstations, or heterogenous combinations of these to complete required computations efficiently.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ibrahim, Khaled Z.; Epifanovsky, Evgeny; Williams, Samuel W.

Coupled-cluster methods provide highly accurate models of molecular structure by explicit numerical calculation of tensors representing the correlation between electrons. These calculations are dominated by a sequence of tensor contractions, motivating the development of numerical libraries for such operations. While based on matrix-matrix multiplication, these libraries are specialized to exploit symmetries in the molecular structure and in electronic interactions, and thus reduce the size of the tensor representation and the complexity of contractions. The resulting algorithms are irregular and their parallelization has been previously achieved via the use of dynamic scheduling or specialized data decompositions. We introduce our efforts tomore » extend the Libtensor framework to work in the distributed memory environment in a scalable and energy efficient manner. We achieve up to 240 speedup compared with the best optimized shared memory implementation. We attain scalability to hundreds of thousands of compute cores on three distributed-memory architectures, (Cray XC30&XC40, BlueGene/Q), and on a heterogeneous GPU-CPU system (Cray XK7). As the bottlenecks shift from being compute-bound DGEMM's to communication-bound collectives as the size of the molecular system scales, we adopt two radically different parallelization approaches for handling load-imbalance. Nevertheless, we preserve a uni ed interface to both programming models to maintain the productivity of computational quantum chemists.« less

A Multi-Level Parallelization Concept for High-Fidelity Multi-Block Solvers

NASA Technical Reports Server (NTRS)

Hatay, Ferhat F.; Jespersen, Dennis C.; Guruswamy, Guru P.; Rizk, Yehia M.; Byun, Chansup; Gee, Ken; VanDalsem, William R. (Technical Monitor)

1997-01-01

The integration of high-fidelity Computational Fluid Dynamics (CFD) analysis tools with the industrial design process benefits greatly from the robust implementations that are transportable across a wide range of computer architectures. In the present work, a hybrid domain-decomposition and parallelization concept was developed and implemented into the widely-used NASA multi-block Computational Fluid Dynamics (CFD) packages implemented in ENSAERO and OVERFLOW. The new parallel solver concept, PENS (Parallel Euler Navier-Stokes Solver), employs both fine and coarse granularity in data partitioning as well as data coalescing to obtain the desired load-balance characteristics on the available computer platforms. This multi-level parallelism implementation itself introduces no changes to the numerical results, hence the original fidelity of the packages are identically preserved. The present implementation uses the Message Passing Interface (MPI) library for interprocessor message passing and memory accessing. By choosing an appropriate combination of the available partitioning and coalescing capabilities only during the execution stage, the PENS solver becomes adaptable to different computer architectures from shared-memory to distributed-memory platforms with varying degrees of parallelism. The PENS implementation on the IBM SP2 distributed memory environment at the NASA Ames Research Center obtains 85 percent scalable parallel performance using fine-grain partitioning of single-block CFD domains using up to 128 wide computational nodes. Multi-block CFD simulations of complete aircraft simulations achieve 75 percent perfect load-balanced executions using data coalescing and the two levels of parallelism. SGI PowerChallenge, SGI Origin 2000, and a cluster of workstations are the other platforms where the robustness of the implementation is tested. The performance behavior on the other computer platforms with a variety of realistic problems will be included as this on-going study progresses.

Scalable Cloning on Large-Scale GPU Platforms with Application to Time-Stepped Simulations on Grids

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yoginath, Srikanth B.; Perumalla, Kalyan S.

Cloning is a technique to efficiently simulate a tree of multiple what-if scenarios that are unraveled during the course of a base simulation. However, cloned execution is highly challenging to realize on large, distributed memory computing platforms, due to the dynamic nature of the computational load across clones, and due to the complex dependencies spanning the clone tree. In this paper, we present the conceptual simulation framework, algorithmic foundations, and runtime interface of CloneX, a new system we designed for scalable simulation cloning. It efficiently and dynamically creates whole logical copies of a dynamic tree of simulations across a largemore » parallel system without full physical duplication of computation and memory. The performance of a prototype implementation executed on up to 1,024 graphical processing units of a supercomputing system has been evaluated with three benchmarks—heat diffusion, forest fire, and disease propagation models—delivering a speed up of over two orders of magnitude compared to replicated runs. Finally, the results demonstrate a significantly faster and scalable way to execute many what-if scenario ensembles of large simulations via cloning using the CloneX interface.« less

Scalable Cloning on Large-Scale GPU Platforms with Application to Time-Stepped Simulations on Grids

DOE PAGES

Yoginath, Srikanth B.; Perumalla, Kalyan S.

2018-01-31

Cloning is a technique to efficiently simulate a tree of multiple what-if scenarios that are unraveled during the course of a base simulation. However, cloned execution is highly challenging to realize on large, distributed memory computing platforms, due to the dynamic nature of the computational load across clones, and due to the complex dependencies spanning the clone tree. In this paper, we present the conceptual simulation framework, algorithmic foundations, and runtime interface of CloneX, a new system we designed for scalable simulation cloning. It efficiently and dynamically creates whole logical copies of a dynamic tree of simulations across a largemore » parallel system without full physical duplication of computation and memory. The performance of a prototype implementation executed on up to 1,024 graphical processing units of a supercomputing system has been evaluated with three benchmarks—heat diffusion, forest fire, and disease propagation models—delivering a speed up of over two orders of magnitude compared to replicated runs. Finally, the results demonstrate a significantly faster and scalable way to execute many what-if scenario ensembles of large simulations via cloning using the CloneX interface.« less

S-HARP: A parallel dynamic spectral partitioner

DOE Office of Scientific and Technical Information (OSTI.GOV)

Sohn, A.; Simon, H.

1998-01-01

Computational science problems with adaptive meshes involve dynamic load balancing when implemented on parallel machines. This dynamic load balancing requires fast partitioning of computational meshes at run time. The authors present in this report a fast parallel dynamic partitioner, called S-HARP. The underlying principles of S-HARP are the fast feature of inertial partitioning and the quality feature of spectral partitioning. S-HARP partitions a graph from scratch, requiring no partition information from previous iterations. Two types of parallelism have been exploited in S-HARP, fine grain loop level parallelism and coarse grain recursive parallelism. The parallel partitioner has been implemented in Messagemore » Passing Interface on Cray T3E and IBM SP2 for portability. Experimental results indicate that S-HARP can partition a mesh of over 100,000 vertices into 256 partitions in 0.2 seconds on a 64 processor Cray T3E. S-HARP is much more scalable than other dynamic partitioners, giving over 15 fold speedup on 64 processors while ParaMeTiS1.0 gives a few fold speedup. Experimental results demonstrate that S-HARP is three to 10 times faster than the dynamic partitioners ParaMeTiS and Jostle on six computational meshes of size over 100,000 vertices.« less

Parallel Climate Data Assimilation PSAS Package

NASA Technical Reports Server (NTRS)

Ding, Hong Q.; Chan, Clara; Gennery, Donald B.; Ferraro, Robert D.

1996-01-01

We have designed and implemented a set of highly efficient and highly scalable algorithms for an unstructured computational package, the PSAS data assimilation package, as demonstrated by detailed performance analysis of systematic runs on up to 512node Intel Paragon. The equation solver achieves a sustained 18 Gflops performance. As the results, we achieved an unprecedented 100-fold solution time reduction on the Intel Paragon parallel platform over the Cray C90. This not only meets and exceeds the DAO time requirements, but also significantly enlarges the window of exploration in climate data assimilations.

Large-scale virtual screening on public cloud resources with Apache Spark.

PubMed

Capuccini, Marco; Ahmed, Laeeq; Schaal, Wesley; Laure, Erwin; Spjuth, Ola

2017-01-01

Structure-based virtual screening is an in-silico method to screen a target receptor against a virtual molecular library. Applying docking-based screening to large molecular libraries can be computationally expensive, however it constitutes a trivially parallelizable task. Most of the available parallel implementations are based on message passing interface, relying on low failure rate hardware and fast network connection. Google's MapReduce revolutionized large-scale analysis, enabling the processing of massive datasets on commodity hardware and cloud resources, providing transparent scalability and fault tolerance at the software level. Open source implementations of MapReduce include Apache Hadoop and the more recent Apache Spark. We developed a method to run existing docking-based screening software on distributed cloud resources, utilizing the MapReduce approach. We benchmarked our method, which is implemented in Apache Spark, docking a publicly available target receptor against [Formula: see text]2.2 M compounds. The performance experiments show a good parallel efficiency (87%) when running in a public cloud environment. Our method enables parallel Structure-based virtual screening on public cloud resources or commodity computer clusters. The degree of scalability that we achieve allows for trying out our method on relatively small libraries first and then to scale to larger libraries. Our implementation is named Spark-VS and it is freely available as open source from GitHub (https://github.com/mcapuccini/spark-vs).Graphical abstract.

Scalable nuclear density functional theory with Sky3D

NASA Astrophysics Data System (ADS)

Afibuzzaman, Md; Schuetrumpf, Bastian; Aktulga, Hasan Metin

2018-02-01

In nuclear astrophysics, quantum simulations of large inhomogeneous dense systems as they appear in the crusts of neutron stars present big challenges. The number of particles in a simulation with periodic boundary conditions is strongly limited due to the immense computational cost of the quantum methods. In this paper, we describe techniques for an efficient and scalable parallel implementation of Sky3D, a nuclear density functional theory solver that operates on an equidistant grid. Presented techniques allow Sky3D to achieve good scaling and high performance on a large number of cores, as demonstrated through detailed performance analysis on a Cray XC40 supercomputer.

Parallel and serial computing tools for testing single-locus and epistatic SNP effects of quantitative traits in genome-wide association studies

PubMed Central

Ma, Li; Runesha, H Birali; Dvorkin, Daniel; Garbe, John R; Da, Yang

2008-01-01

Background Genome-wide association studies (GWAS) using single nucleotide polymorphism (SNP) markers provide opportunities to detect epistatic SNPs associated with quantitative traits and to detect the exact mode of an epistasis effect. Computational difficulty is the main bottleneck for epistasis testing in large scale GWAS. Results The EPISNPmpi and EPISNP computer programs were developed for testing single-locus and epistatic SNP effects on quantitative traits in GWAS, including tests of three single-locus effects for each SNP (SNP genotypic effect, additive and dominance effects) and five epistasis effects for each pair of SNPs (two-locus interaction, additive × additive, additive × dominance, dominance × additive, and dominance × dominance) based on the extended Kempthorne model. EPISNPmpi is the parallel computing program for epistasis testing in large scale GWAS and achieved excellent scalability for large scale analysis and portability for various parallel computing platforms. EPISNP is the serial computing program based on the EPISNPmpi code for epistasis testing in small scale GWAS using commonly available operating systems and computer hardware. Three serial computing utility programs were developed for graphical viewing of test results and epistasis networks, and for estimating CPU time and disk space requirements. Conclusion The EPISNPmpi parallel computing program provides an effective computing tool for epistasis testing in large scale GWAS, and the epiSNP serial computing programs are convenient tools for epistasis analysis in small scale GWAS using commonly available computer hardware. PMID:18644146

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chen, Chao; Pouransari, Hadi; Rajamanickam, Sivasankaran

We present a parallel hierarchical solver for general sparse linear systems on distributed-memory machines. For large-scale problems, this fully algebraic algorithm is faster and more memory-efficient than sparse direct solvers because it exploits the low-rank structure of fill-in blocks. Depending on the accuracy of low-rank approximations, the hierarchical solver can be used either as a direct solver or as a preconditioner. The parallel algorithm is based on data decomposition and requires only local communication for updating boundary data on every processor. Moreover, the computation-to-communication ratio of the parallel algorithm is approximately the volume-to-surface-area ratio of the subdomain owned by everymore » processor. We also provide various numerical results to demonstrate the versatility and scalability of the parallel algorithm.« less

Visual analysis of inter-process communication for large-scale parallel computing.

PubMed

Muelder, Chris; Gygi, Francois; Ma, Kwan-Liu

2009-01-01

In serial computation, program profiling is often helpful for optimization of key sections of code. When moving to parallel computation, not only does the code execution need to be considered but also communication between the different processes which can induce delays that are detrimental to performance. As the number of processes increases, so does the impact of the communication delays on performance. For large-scale parallel applications, it is critical to understand how the communication impacts performance in order to make the code more efficient. There are several tools available for visualizing program execution and communications on parallel systems. These tools generally provide either views which statistically summarize the entire program execution or process-centric views. However, process-centric visualizations do not scale well as the number of processes gets very large. In particular, the most common representation of parallel processes is a Gantt char t with a row for each process. As the number of processes increases, these charts can become difficult to work with and can even exceed screen resolution. We propose a new visualization approach that affords more scalability and then demonstrate it on systems running with up to 16,384 processes.

SciSpark: Highly Interactive and Scalable Model Evaluation and Climate Metrics

NASA Astrophysics Data System (ADS)

Wilson, B. D.; Mattmann, C. A.; Waliser, D. E.; Kim, J.; Loikith, P.; Lee, H.; McGibbney, L. J.; Whitehall, K. D.

2014-12-01

Remote sensing data and climate model output are multi-dimensional arrays of massive sizes locked away in heterogeneous file formats (HDF5/4, NetCDF 3/4) and metadata models (HDF-EOS, CF) making it difficult to perform multi-stage, iterative science processing since each stage requires writing and reading data to and from disk. We are developing a lightning fast Big Data technology called SciSpark based on ApacheTM Spark. Spark implements the map-reduce paradigm for parallel computing on a cluster, but emphasizes in-memory computation, "spilling" to disk only as needed, and so outperforms the disk-based ApacheTM Hadoop by 100x in memory and by 10x on disk, and makes iterative algorithms feasible. SciSpark will enable scalable model evaluation by executing large-scale comparisons of A-Train satellite observations to model grids on a cluster of 100 to 1000 compute nodes. This 2nd generation capability for NASA's Regional Climate Model Evaluation System (RCMES) will compute simple climate metrics at interactive speeds, and extend to quite sophisticated iterative algorithms such as machine-learning (ML) based clustering of temperature PDFs, and even graph-based algorithms for searching for Mesocale Convective Complexes. The goals of SciSpark are to: (1) Decrease the time to compute comparison statistics and plots from minutes to seconds; (2) Allow for interactive exploration of time-series properties over seasons and years; (3) Decrease the time for satellite data ingestion into RCMES to hours; (4) Allow for Level-2 comparisons with higher-order statistics or PDF's in minutes to hours; and (5) Move RCMES into a near real time decision-making platform. We will report on: the architecture and design of SciSpark, our efforts to integrate climate science algorithms in Python and Scala, parallel ingest and partitioning (sharding) of A-Train satellite observations from HDF files and model grids from netCDF files, first parallel runs to compute comparison statistics and PDF's, and first metrics quantifying parallel speedups and memory & disk usage.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Janjusic, Tommy; Kartsaklis, Christos

Memory scalability is an enduring problem and bottleneck that plagues many parallel codes. Parallel codes designed for High Performance Systems are typically designed over the span of several, and in some instances 10+, years. As a result, optimization practices which were appropriate for earlier systems may no longer be valid and thus require careful optimization consideration. Specifically, parallel codes whose memory footprint is a function of their scalability must be carefully considered for future exa-scale systems. In this paper we present a methodology and tool to study the memory scalability of parallel codes. Using our methodology we evaluate an applicationmore » s memory footprint as a function of scalability, which we coined memory efficiency, and describe our results. In particular, using our in-house tools we can pinpoint the specific application components which contribute to the application s overall memory foot-print (application data- structures, libraries, etc.).« less

Topical perspective on massive threading and parallelism.

PubMed

Farber, Robert M

2011-09-01

Unquestionably computer architectures have undergone a recent and noteworthy paradigm shift that now delivers multi- and many-core systems with tens to many thousands of concurrent hardware processing elements per workstation or supercomputer node. GPGPU (General Purpose Graphics Processor Unit) technology in particular has attracted significant attention as new software development capabilities, namely CUDA (Compute Unified Device Architecture) and OpenCL™, have made it possible for students as well as small and large research organizations to achieve excellent speedup for many applications over more conventional computing architectures. The current scientific literature reflects this shift with numerous examples of GPGPU applications that have achieved one, two, and in some special cases, three-orders of magnitude increased computational performance through the use of massive threading to exploit parallelism. Multi-core architectures are also evolving quickly to exploit both massive-threading and massive-parallelism such as the 1.3 million threads Blue Waters supercomputer. The challenge confronting scientists in planning future experimental and theoretical research efforts--be they individual efforts with one computer or collaborative efforts proposing to use the largest supercomputers in the world is how to capitalize on these new massively threaded computational architectures--especially as not all computational problems will scale to massive parallelism. In particular, the costs associated with restructuring software (and potentially redesigning algorithms) to exploit the parallelism of these multi- and many-threaded machines must be considered along with application scalability and lifespan. This perspective is an overview of the current state of threading and parallelize with some insight into the future. Published by Elsevier Inc.

Scalable descriptive and correlative statistics with Titan.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Thompson, David C.; Pebay, Philippe Pierre

This report summarizes the existing statistical engines in VTK/Titan and presents the parallel versions thereof which have already been implemented. The ease of use of these parallel engines is illustrated by the means of C++ code snippets. Furthermore, this report justifies the design of these engines with parallel scalability in mind; then, this theoretical property is verified with test runs that demonstrate optimal parallel speed-up with up to 200 processors.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ali, Amjad Majid; Albert, Don; Andersson, Par

SLURM is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for large and small computer clusters. As a cluster resource manager, SLURM has three key functions. First, it allocates exclusive and/or non-exclusive access to resources (compute nodes) to users for some duration of time so they can perform work. Second, it provides a framework for starting, executing, and monitoring work 9normally a parallel job) on the set of allocated nodes. Finally, it arbitrates conflicting requests for resources by managing a queue of pending work.

SciSpark: Highly Interactive and Scalable Model Evaluation and Climate Metrics

NASA Astrophysics Data System (ADS)

Wilson, B. D.; Palamuttam, R. S.; Mogrovejo, R. M.; Whitehall, K. D.; Mattmann, C. A.; Verma, R.; Waliser, D. E.; Lee, H.

2015-12-01

Remote sensing data and climate model output are multi-dimensional arrays of massive sizes locked away in heterogeneous file formats (HDF5/4, NetCDF 3/4) and metadata models (HDF-EOS, CF) making it difficult to perform multi-stage, iterative science processing since each stage requires writing and reading data to and from disk. We are developing a lightning fast Big Data technology called SciSpark based on ApacheTM Spark under a NASA AIST grant (PI Mattmann). Spark implements the map-reduce paradigm for parallel computing on a cluster, but emphasizes in-memory computation, "spilling" to disk only as needed, and so outperforms the disk-based ApacheTM Hadoop by 100x in memory and by 10x on disk. SciSpark will enable scalable model evaluation by executing large-scale comparisons of A-Train satellite observations to model grids on a cluster of 10 to 1000 compute nodes. This 2nd generation capability for NASA's Regional Climate Model Evaluation System (RCMES) will compute simple climate metrics at interactive speeds, and extend to quite sophisticated iterative algorithms such as machine-learning based clustering of temperature PDFs, and even graph-based algorithms for searching for Mesocale Convective Complexes. We have implemented a parallel data ingest capability in which the user specifies desired variables (arrays) as several time-sorted lists of URL's (i.e. using OPeNDAP model.nc?varname, or local files). The specified variables are partitioned by time/space and then each Spark node pulls its bundle of arrays into memory to begin a computation pipeline. We also investigated the performance of several N-dim. array libraries (scala breeze, java jblas & netlib-java, and ND4J). We are currently developing science codes using ND4J and studying memory behavior on the JVM. On the pyspark side, many of our science codes already use the numpy and SciPy ecosystems. The talk will cover: the architecture of SciSpark, the design of the scientific RDD (sRDD) data structure, our efforts to integrate climate science algorithms in Python and Scala, parallel ingest and partitioning of A-Train satellite observations from HDF files and model grids from netCDF files, first parallel runs to compute comparison statistics and PDF's, and first metrics quantifying parallel speedups and memory & disk usage.

Preparing for Exascale: Towards convection-permitting, global atmospheric simulations with the Model for Prediction Across Scales (MPAS)

NASA Astrophysics Data System (ADS)

Heinzeller, Dominikus; Duda, Michael G.; Kunstmann, Harald

2017-04-01

With strong financial and political support from national and international initiatives, exascale computing is projected for the end of this decade. Energy requirements and physical limitations imply the use of accelerators and the scaling out to orders of magnitudes larger numbers of cores then today to achieve this milestone. In order to fully exploit the capabilities of these Exascale computing systems, existing applications need to undergo significant development. The Model for Prediction Across Scales (MPAS) is a novel set of Earth system simulation components and consists of an atmospheric core, an ocean core, a land-ice core and a sea-ice core. Its distinct features are the use of unstructured Voronoi meshes and C-grid discretisation to address shortcomings of global models on regular grids and the use of limited area models nested in a forcing data set, with respect to parallel scalability, numerical accuracy and physical consistency. Here, we present work towards the application of the atmospheric core (MPAS-A) on current and future high performance computing systems for problems at extreme scale. In particular, we address the issue of massively parallel I/O by extending the model to support the highly scalable SIONlib library. Using global uniform meshes with a convection-permitting resolution of 2-3km, we demonstrate the ability of MPAS-A to scale out to half a million cores while maintaining a high parallel efficiency. We also demonstrate the potential benefit of a hybrid parallelisation of the code (MPI/OpenMP) on the latest generation of Intel's Many Integrated Core Architecture, the Intel Xeon Phi Knights Landing.

Cheetah: A Framework for Scalable Hierarchical Collective Operations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Graham, Richard L; Gorentla Venkata, Manjunath; Ladd, Joshua S

2011-01-01

Collective communication operations, used by many scientific applications, tend to limit overall parallel application performance and scalability. Computer systems are becoming more heterogeneous with increasing node and core-per-node counts. Also, a growing number of data-access mechanisms, of varying characteristics, are supported within a single computer system. We describe a new hierarchical collective communication framework that takes advantage of hardware-specific data-access mechanisms. It is flexible, with run-time hierarchy specification, and sharing of collective communication primitives between collective algorithms. Data buffers are shared between levels in the hierarchy reducing collective communication management overhead. We have implemented several versions of the Message Passingmore » Interface (MPI) collective operations, MPI Barrier() and MPI Bcast(), and run experiments using up to 49, 152 processes on a Cray XT5, and a small InfiniBand based cluster. At 49, 152 processes our barrier implementation outperforms the optimized native implementation by 75%. 32 Byte and one Mega-Byte broadcasts outperform it by 62% and 11%, respectively, with better scalability characteristics. Improvements relative to the default Open MPI implementation are much larger.« less

Parallelization of an Object-Oriented Unstructured Aeroacoustics Solver

NASA Technical Reports Server (NTRS)

Baggag, Abdelkader; Atkins, Harold; Oezturan, Can; Keyes, David

1999-01-01

A computational aeroacoustics code based on the discontinuous Galerkin method is ported to several parallel platforms using MPI. The discontinuous Galerkin method is a compact high-order method that retains its accuracy and robustness on non-smooth unstructured meshes. In its semi-discrete form, the discontinuous Galerkin method can be combined with explicit time marching methods making it well suited to time accurate computations. The compact nature of the discontinuous Galerkin method also makes it well suited for distributed memory parallel platforms. The original serial code was written using an object-oriented approach and was previously optimized for cache-based machines. The port to parallel platforms was achieved simply by treating partition boundaries as a type of boundary condition. Code modifications were minimal because boundary conditions were abstractions in the original program. Scalability results are presented for the SCI Origin, IBM SP2, and clusters of SGI and Sun workstations. Slightly superlinear speedup is achieved on a fixed-size problem on the Origin, due to cache effects.

Performance assessment of KORAT-3D on the ANL IBM-SP computer

DOE Office of Scientific and Technical Information (OSTI.GOV)

Alexeyev, A.V.; Zvenigorodskaya, O.A.; Shagaliev, R.M.

1999-09-01

The TENAR code is currently being developed at the Russian Federal Nuclear Center (VNIIEF) as a coupled dynamics code for the simulation of transients in VVER and RBMK systems and other nuclear systems. The neutronic module in this code system is KORAT-3D. This module is also one of the most computationally intensive components of the code system. A parallel version of KORAT-3D has been implemented to achieve the goal of obtaining transient solutions in reasonable computational time, particularly for RBMK calculations that involve the application of >100,000 nodes. An evaluation of the KORAT-3D code performance was recently undertaken on themore » Argonne National Laboratory (ANL) IBM ScalablePower (SP) parallel computer located in the Mathematics and Computer Science Division of ANL. At the time of the study, the ANL IBM-SP computer had 80 processors. This study was conducted under the auspices of a technical staff exchange program sponsored by the International Nuclear Safety Center (INSC).« less

Use of Massive Parallel Computing Libraries in the Context of Global Gravity Field Determination from Satellite Data

NASA Astrophysics Data System (ADS)

Brockmann, J. M.; Schuh, W.-D.

2011-07-01

The estimation of the global Earth's gravity field parametrized as a finite spherical harmonic series is computationally demanding. The computational effort depends on the one hand on the maximal resolution of the spherical harmonic expansion (i.e. the number of parameters to be estimated) and on the other hand on the number of observations (which are several millions for e.g. observations from the GOCE satellite missions). To circumvent these restrictions, a massive parallel software based on high-performance computing (HPC) libraries as ScaLAPACK, PBLAS and BLACS was designed in the context of GOCE HPF WP6000 and the GOCO consortium. A prerequisite for the use of these libraries is that all matrices are block-cyclic distributed on a processor grid comprised by a large number of (distributed memory) computers. Using this set of standard HPC libraries has the benefit that once the matrices are distributed across the computer cluster, a huge set of efficient and highly scalable linear algebra operations can be used.

Zonal methods for the parallel execution of range-limited N-body simulations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Bowers, Kevin J.; Dror, Ron O.; Shaw, David E.

2007-01-20

Particle simulations in fields ranging from biochemistry to astrophysics require the evaluation of interactions between all pairs of particles separated by less than some fixed interaction radius. The applicability of such simulations is often limited by the time required for calculation, but the use of massive parallelism to accelerate these computations is typically limited by inter-processor communication requirements. Recently, Snir [M. Snir, A note on N-body computations with cutoffs, Theor. Comput. Syst. 37 (2004) 295-318] and Shaw [D.E. Shaw, A fast, scalable method for the parallel evaluation of distance-limited pairwise particle interactions, J. Comput. Chem. 26 (2005) 1318-1328] independently introducedmore » two distinct methods that offer asymptotic reductions in the amount of data transferred between processors. In the present paper, we show that these schemes represent special cases of a more general class of methods, and introduce several new algorithms in this class that offer practical advantages over all previously described methods for a wide range of problem parameters. We also show that several of these algorithms approach an approximate lower bound on inter-processor data transfer.« less

PROTO-PLASM: parallel language for adaptive and scalable modelling of biosystems.

PubMed

Bajaj, Chandrajit; DiCarlo, Antonio; Paoluzzi, Alberto

2008-09-13

This paper discusses the design goals and the first developments of PROTO-PLASM, a novel computational environment to produce libraries of executable, combinable and customizable computer models of natural and synthetic biosystems, aiming to provide a supporting framework for predictive understanding of structure and behaviour through multiscale geometric modelling and multiphysics simulations. Admittedly, the PROTO-PLASM platform is still in its infancy. Its computational framework--language, model library, integrated development environment and parallel engine--intends to provide patient-specific computational modelling and simulation of organs and biosystem, exploiting novel functionalities resulting from the symbolic combination of parametrized models of parts at various scales. PROTO-PLASM may define the model equations, but it is currently focused on the symbolic description of model geometry and on the parallel support of simulations. Conversely, CellML and SBML could be viewed as defining the behavioural functions (the model equations) to be used within a PROTO-PLASM program. Here we exemplify the basic functionalities of PROTO-PLASM, by constructing a schematic heart model. We also discuss multiscale issues with reference to the geometric and physical modelling of neuromuscular junctions.

Proto-Plasm: parallel language for adaptive and scalable modelling of biosystems

PubMed Central

Bajaj, Chandrajit; DiCarlo, Antonio; Paoluzzi, Alberto

2008-01-01

This paper discusses the design goals and the first developments of Proto-Plasm, a novel computational environment to produce libraries of executable, combinable and customizable computer models of natural and synthetic biosystems, aiming to provide a supporting framework for predictive understanding of structure and behaviour through multiscale geometric modelling and multiphysics simulations. Admittedly, the Proto-Plasm platform is still in its infancy. Its computational framework—language, model library, integrated development environment and parallel engine—intends to provide patient-specific computational modelling and simulation of organs and biosystem, exploiting novel functionalities resulting from the symbolic combination of parametrized models of parts at various scales. Proto-Plasm may define the model equations, but it is currently focused on the symbolic description of model geometry and on the parallel support of simulations. Conversely, CellML and SBML could be viewed as defining the behavioural functions (the model equations) to be used within a Proto-Plasm program. Here we exemplify the basic functionalities of Proto-Plasm, by constructing a schematic heart model. We also discuss multiscale issues with reference to the geometric and physical modelling of neuromuscular junctions. PMID:18559320

Parallel Dynamics Simulation Using a Krylov-Schwarz Linear Solution Scheme

DOE PAGES

Abhyankar, Shrirang; Constantinescu, Emil M.; Smith, Barry F.; ...

2016-11-07

Fast dynamics simulation of large-scale power systems is a computational challenge because of the need to solve a large set of stiff, nonlinear differential-algebraic equations at every time step. The main bottleneck in dynamic simulations is the solution of a linear system during each nonlinear iteration of Newton’s method. In this paper, we present a parallel Krylov- Schwarz linear solution scheme that uses the Krylov subspacebased iterative linear solver GMRES with an overlapping restricted additive Schwarz preconditioner. As a result, performance tests of the proposed Krylov-Schwarz scheme for several large test cases ranging from 2,000 to 20,000 buses, including amore » real utility network, show good scalability on different computing architectures.« less

Parallel Dynamics Simulation Using a Krylov-Schwarz Linear Solution Scheme

DOE Office of Scientific and Technical Information (OSTI.GOV)

Abhyankar, Shrirang; Constantinescu, Emil M.; Smith, Barry F.

Fast dynamics simulation of large-scale power systems is a computational challenge because of the need to solve a large set of stiff, nonlinear differential-algebraic equations at every time step. The main bottleneck in dynamic simulations is the solution of a linear system during each nonlinear iteration of Newton’s method. In this paper, we present a parallel Krylov- Schwarz linear solution scheme that uses the Krylov subspacebased iterative linear solver GMRES with an overlapping restricted additive Schwarz preconditioner. As a result, performance tests of the proposed Krylov-Schwarz scheme for several large test cases ranging from 2,000 to 20,000 buses, including amore » real utility network, show good scalability on different computing architectures.« less

DOE Office of Scientific and Technical Information (OSTI.GOV)

Rajbhandari, Samyam; NIkam, Akshay; Lai, Pai-Wei

Tensor contractions represent the most compute-intensive core kernels in ab initio computational quantum chemistry and nuclear physics. Symmetries in these tensor contractions makes them difficult to load balance and scale to large distributed systems. In this paper, we develop an efficient and scalable algorithm to contract symmetric tensors. We introduce a novel approach that avoids data redistribution in contracting symmetric tensors while also avoiding redundant storage and maintaining load balance. We present experimental results on two parallel supercomputers for several symmetric contractions that appear in the CCSD quantum chemistry method. We also present a novel approach to tensor redistribution thatmore » can take advantage of parallel hyperplanes when the initial distribution has replicated dimensions, and use collective broadcast when the final distribution has replicated dimensions, making the algorithm very efficient.« less

Large-scale Parallel Unstructured Mesh Computations for 3D High-lift Analysis

NASA Technical Reports Server (NTRS)

Mavriplis, Dimitri J.; Pirzadeh, S.

1999-01-01

A complete "geometry to drag-polar" analysis capability for the three-dimensional high-lift configurations is described. The approach is based on the use of unstructured meshes in order to enable rapid turnaround for complicated geometries that arise in high-lift configurations. Special attention is devoted to creating a capability for enabling analyses on highly resolved grids. Unstructured meshes of several million vertices are initially generated on a work-station, and subsequently refined on a supercomputer. The flow is solved on these refined meshes on large parallel computers using an unstructured agglomeration multigrid algorithm. Good prediction of lift and drag throughout the range of incidences is demonstrated on a transport take-off configuration using up to 24.7 million grid points. The feasibility of using this approach in a production environment on existing parallel machines is demonstrated, as well as the scalability of the solver on machines using up to 1450 processors.

Automation of Data Traffic Control on DSM Architecture

NASA Technical Reports Server (NTRS)

Frumkin, Michael; Jin, Hao-Qiang; Yan, Jerry

2001-01-01

The design of distributed shared memory (DSM) computers liberates users from the duty to distribute data across processors and allows for the incremental development of parallel programs using, for example, OpenMP or Java threads. DSM architecture greatly simplifies the development of parallel programs having good performance on a few processors. However, to achieve a good program scalability on DSM computers requires that the user understand data flow in the application and use various techniques to avoid data traffic congestions. In this paper we discuss a number of such techniques, including data blocking, data placement, data transposition and page size control and evaluate their efficiency on the NAS (NASA Advanced Supercomputing) Parallel Benchmarks. We also present a tool which automates the detection of constructs causing data congestions in Fortran array oriented codes and advises the user on code transformations for improving data traffic in the application.

PRIMA-X Final Report

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lorenz, Daniel; Wolf, Felix

2016-02-17

The PRIMA-X (Performance Retargeting of Instrumentation, Measurement, and Analysis Technologies for Exascale Computing) project is the successor of the DOE PRIMA (Performance Refactoring of Instrumentation, Measurement, and Analysis Technologies for Petascale Computing) project, which addressed the challenge of creating a core measurement infrastructure that would serve as a common platform for both integrating leading parallel performance systems (notably TAU and Scalasca) and developing next-generation scalable performance tools. The PRIMA-X project shifts the focus away from refactorization of robust performance tools towards a re-targeting of the parallel performance measurement and analysis architecture for extreme scales. The massive concurrency, asynchronous execution dynamics,more » hardware heterogeneity, and multi-objective prerequisites (performance, power, resilience) that identify exascale systems introduce fundamental constraints on the ability to carry forward existing performance methodologies. In particular, there must be a deemphasis of per-thread observation techniques to significantly reduce the otherwise unsustainable flood of redundant performance data. Instead, it will be necessary to assimilate multi-level resource observations into macroscopic performance views, from which resilient performance metrics can be attributed to the computational features of the application. This requires a scalable framework for node-level and system-wide monitoring and runtime analyses of dynamic performance information. Also, the interest in optimizing parallelism parameters with respect to performance and energy drives the integration of tool capabilities in the exascale environment further. Initially, PRIMA-X was a collaborative project between the University of Oregon (lead institution) and the German Research School for Simulation Sciences (GRS). Because Prof. Wolf, the PI at GRS, accepted a position as full professor at Technische Universität Darmstadt (TU Darmstadt) starting February 1st, 2015, the project ended at GRS on January 31st, 2015. This report reflects the work accomplished at GRS until then. The work of GRS is expected to be continued at TU Darmstadt. The first main accomplishment of GRS is the design of different thread-level aggregation techniques. We created a prototype capable of aggregating the thread-level information in performance profiles using these techniques. The next step will be the integration of the most promising techniques into the Score-P measurement system and their evaluation. The second main accomplishment is a substantial increase of Score-P’s scalability, achieved by improving the design of the system-tree representation in Score-P’s profile format. We developed a new representation and a distributed algorithm to create the scalable system tree representation. Finally, we developed a lightweight approach to MPI wait-state profiling. Former algorithms either needed piggy-backing, which can cause significant runtime overhead, or tracing, which comes with its own set of scaling challenges. Our approach works with local data only and, thus, is scalable and has very little overhead.« less

Final Scientific Report: A Scalable Development Environment for Peta-Scale Computing

DOE Office of Scientific and Technical Information (OSTI.GOV)

Karbach, Carsten; Frings, Wolfgang

2013-02-22

This document is the final scientific report of the project DE-SC000120 (A scalable Development Environment for Peta-Scale Computing). The objective of this project is the extension of the Parallel Tools Platform (PTP) for applying it to peta-scale systems. PTP is an integrated development environment for parallel applications. It comprises code analysis, performance tuning, parallel debugging and system monitoring. The contribution of the Juelich Supercomputing Centre (JSC) aims to provide a scalable solution for system monitoring of supercomputers. This includes the development of a new communication protocol for exchanging status data between the target remote system and the client running PTP.more » The communication has to work for high latency. PTP needs to be implemented robustly and should hide the complexity of the supercomputer's architecture in order to provide a transparent access to various remote systems via a uniform user interface. This simplifies the porting of applications to different systems, because PTP functions as abstraction layer between parallel application developer and compute resources. The common requirement for all PTP components is that they have to interact with the remote supercomputer. E.g. applications are built remotely and performance tools are attached to job submissions and their output data resides on the remote system. Status data has to be collected by evaluating outputs of the remote job scheduler and the parallel debugger needs to control an application executed on the supercomputer. The challenge is to provide this functionality for peta-scale systems in real-time. The client server architecture of the established monitoring application LLview, developed by the JSC, can be applied to PTP's system monitoring. LLview provides a well-arranged overview of the supercomputer's current status. A set of statistics, a list of running and queued jobs as well as a node display mapping running jobs to their compute resources form the user display of LLview. These monitoring features have to be integrated into the development environment. Besides showing the current status PTP's monitoring also needs to allow for submitting and canceling user jobs. Monitoring peta-scale systems especially deals with presenting the large amount of status data in a useful manner. Users require to select arbitrary levels of detail. The monitoring views have to provide a quick overview of the system state, but also need to allow for zooming into specific parts of the system, into which the user is interested in. At present, the major batch systems running on supercomputers are PBS, TORQUE, ALPS and LoadLeveler, which have to be supported by both the monitoring and the job controlling component. Finally, PTP needs to be designed as generic as possible, so that it can be extended for future batch systems.« less

Parallel Simulation of Three-Dimensional Free Surface Fluid Flow Problems

DOE Office of Scientific and Technical Information (OSTI.GOV)

BAER,THOMAS A.; SACKINGER,PHILIP A.; SUBIA,SAMUEL R.

1999-10-14

Simulation of viscous three-dimensional fluid flow typically involves a large number of unknowns. When free surfaces are included, the number of unknowns increases dramatically. Consequently, this class of problem is an obvious application of parallel high performance computing. We describe parallel computation of viscous, incompressible, free surface, Newtonian fluid flow problems that include dynamic contact fines. The Galerkin finite element method was used to discretize the fully-coupled governing conservation equations and a ''pseudo-solid'' mesh mapping approach was used to determine the shape of the free surface. In this approach, the finite element mesh is allowed to deform to satisfy quasi-staticmore » solid mechanics equations subject to geometric or kinematic constraints on the boundaries. As a result, nodal displacements must be included in the set of unknowns. Other issues discussed are the proper constraints appearing along the dynamic contact line in three dimensions. Issues affecting efficient parallel simulations include problem decomposition to equally distribute computational work among a SPMD computer and determination of robust, scalable preconditioners for the distributed matrix systems that must be solved. Solution continuation strategies important for serial simulations have an enhanced relevance in a parallel coquting environment due to the difficulty of solving large scale systems. Parallel computations will be demonstrated on an example taken from the coating flow industry: flow in the vicinity of a slot coater edge. This is a three dimensional free surface problem possessing a contact line that advances at the web speed in one region but transitions to static behavior in another region. As such, a significant fraction of the computational time is devoted to processing boundary data. Discussion focuses on parallel speed ups for fixed problem size, a class of problems of immediate practical importance.« less

An efficient parallel algorithm: Poststack and prestack Kirchhoff 3D depth migration using flexi-depth iterations

NASA Astrophysics Data System (ADS)

Rastogi, Richa; Srivastava, Abhishek; Khonde, Kiran; Sirasala, Kirannmayi M.; Londhe, Ashutosh; Chavhan, Hitesh

2015-07-01

This paper presents an efficient parallel 3D Kirchhoff depth migration algorithm suitable for current class of multicore architecture. The fundamental Kirchhoff depth migration algorithm exhibits inherent parallelism however, when it comes to 3D data migration, as the data size increases the resource requirement of the algorithm also increases. This challenges its practical implementation even on current generation high performance computing systems. Therefore a smart parallelization approach is essential to handle 3D data for migration. The most compute intensive part of Kirchhoff depth migration algorithm is the calculation of traveltime tables due to its resource requirements such as memory/storage and I/O. In the current research work, we target this area and develop a competent parallel algorithm for post and prestack 3D Kirchhoff depth migration, using hybrid MPI+OpenMP programming techniques. We introduce a concept of flexi-depth iterations while depth migrating data in parallel imaging space, using optimized traveltime table computations. This concept provides flexibility to the algorithm by migrating data in a number of depth iterations, which depends upon the available node memory and the size of data to be migrated during runtime. Furthermore, it minimizes the requirements of storage, I/O and inter-node communication, thus making it advantageous over the conventional parallelization approaches. The developed parallel algorithm is demonstrated and analysed on Yuva II, a PARAM series of supercomputers. Optimization, performance and scalability experiment results along with the migration outcome show the effectiveness of the parallel algorithm.

Reversible Parallel Discrete-Event Execution of Large-scale Epidemic Outbreak Models

DOE Office of Scientific and Technical Information (OSTI.GOV)

Perumalla, Kalyan S; Seal, Sudip K

2010-01-01

The spatial scale, runtime speed and behavioral detail of epidemic outbreak simulations together require the use of large-scale parallel processing. In this paper, an optimistic parallel discrete event execution of a reaction-diffusion simulation model of epidemic outbreaks is presented, with an implementation over themore » $$\\mu$$sik simulator. Rollback support is achieved with the development of a novel reversible model that combines reverse computation with a small amount of incremental state saving. Parallel speedup and other runtime performance metrics of the simulation are tested on a small (8,192-core) Blue Gene / P system, while scalability is demonstrated on 65,536 cores of a large Cray XT5 system. Scenarios representing large population sizes (up to several hundred million individuals in the largest case) are exercised.« less

Parallel-aware, dedicated job co-scheduling within/across symmetric multiprocessing nodes

DOEpatents

Jones, Terry R.; Watson, Pythagoras C.; Tuel, William; Brenner, Larry; ,Caffrey, Patrick; Fier, Jeffrey

2010-10-05

In a parallel computing environment comprising a network of SMP nodes each having at least one processor, a parallel-aware co-scheduling method and system for improving the performance and scalability of a dedicated parallel job having synchronizing collective operations. The method and system uses a global co-scheduler and an operating system kernel dispatcher adapted to coordinate interfering system and daemon activities on a node and across nodes to promote intra-node and inter-node overlap of said interfering system and daemon activities as well as intra-node and inter-node overlap of said synchronizing collective operations. In this manner, the impact of random short-lived interruptions, such as timer-decrement processing and periodic daemon activity, on synchronizing collective operations is minimized on large processor-count SPMD bulk-synchronous programming styles.

Department of Defense High Performance Computing Modernization Program. 2006 Annual Report

DTIC Science & Technology

2007-03-01

Department. We successfully completed several software development projects that introduced parallel, scalable production software now in use across the...imagined. They are developing and deploying weather and ocean models that allow our soldiers, sailors, marines and airmen to plan missions more effectively...and to navigate adverse environments safely. They are modeling molecular interactions leading to the development of higher energy fuels, munitions

A Scalable Implementation of Van der Waals Density Functionals

NASA Astrophysics Data System (ADS)

Wu, Jun; Gygi, Francois

2010-03-01

Recently developed Van der Waals density functionals[1] offer the promise to account for weak intermolecular interactions that are not described accurately by local exchange-correlation density functionals. In spite of recent progress [2], the computational cost of such calculations remains high. We present a scalable parallel implementation of the functional proposed by Dion et al.[1]. The method is implemented in the Qbox first-principles simulation code (http://eslab.ucdavis.edu/software/qbox). Application to large molecular systems will be presented. [4pt] [1] M. Dion et al. Phys. Rev. Lett. 92, 246401 (2004).[0pt] [2] G. Roman-Perez and J. M. Soler, Phys. Rev. Lett. 103, 096102 (2009).

Π4U: A high performance computing framework for Bayesian uncertainty quantification of complex models

NASA Astrophysics Data System (ADS)

Hadjidoukas, P. E.; Angelikopoulos, P.; Papadimitriou, C.; Koumoutsakos, P.

2015-03-01

We present Π4U, an extensible framework, for non-intrusive Bayesian Uncertainty Quantification and Propagation (UQ+P) of complex and computationally demanding physical models, that can exploit massively parallel computer architectures. The framework incorporates Laplace asymptotic approximations as well as stochastic algorithms, along with distributed numerical differentiation and task-based parallelism for heterogeneous clusters. Sampling is based on the Transitional Markov Chain Monte Carlo (TMCMC) algorithm and its variants. The optimization tasks associated with the asymptotic approximations are treated via the Covariance Matrix Adaptation Evolution Strategy (CMA-ES). A modified subset simulation method is used for posterior reliability measurements of rare events. The framework accommodates scheduling of multiple physical model evaluations based on an adaptive load balancing library and shows excellent scalability. In addition to the software framework, we also provide guidelines as to the applicability and efficiency of Bayesian tools when applied to computationally demanding physical models. Theoretical and computational developments are demonstrated with applications drawn from molecular dynamics, structural dynamics and granular flow.

The GBS code for tokamak scrape-off layer simulations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Halpern, F.D., E-mail: federico.halpern@epfl.ch; Ricci, P.; Jolliet, S.

2016-06-15

We describe a new version of GBS, a 3D global, flux-driven plasma turbulence code to simulate the turbulent dynamics in the tokamak scrape-off layer (SOL), superseding the code presented by Ricci et al. (2012) [14]. The present work is driven by the objective of studying SOL turbulent dynamics in medium size tokamaks and beyond with a high-fidelity physics model. We emphasize an intertwining framework of improved physics models and the computational improvements that allow them. The model extensions include neutral atom physics, finite ion temperature, the addition of a closed field line region, and a non-Boussinesq treatment of the polarizationmore » drift. GBS has been completely refactored with the introduction of a 3-D Cartesian communicator and a scalable parallel multigrid solver. We report dramatically enhanced parallel scalability, with the possibility of treating electromagnetic fluctuations very efficiently. The method of manufactured solutions as a verification process has been carried out for this new code version, demonstrating the correct implementation of the physical model.« less

OWL: A scalable Monte Carlo simulation suite for finite-temperature study of materials

NASA Astrophysics Data System (ADS)

Li, Ying Wai; Yuk, Simuck F.; Cooper, Valentino R.; Eisenbach, Markus; Odbadrakh, Khorgolkhuu

The OWL suite is a simulation package for performing large-scale Monte Carlo simulations. Its object-oriented, modular design enables it to interface with various external packages for energy evaluations. It is therefore applicable to study the finite-temperature properties for a wide range of systems: from simple classical spin models to materials where the energy is evaluated by ab initio methods. This scheme not only allows for the study of thermodynamic properties based on first-principles statistical mechanics, it also provides a means for massive, multi-level parallelism to fully exploit the capacity of modern heterogeneous computer architectures. We will demonstrate how improved strong and weak scaling is achieved by employing novel, parallel and scalable Monte Carlo algorithms, as well as the applications of OWL to a few selected frontier materials research problems. This research was supported by the Office of Science of the Department of Energy under contract DE-AC05-00OR22725.

Final report for “Extreme-scale Algorithms and Solver Resilience”

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gropp, William Douglas

2017-06-30

This is a joint project with principal investigators at Oak Ridge National Laboratory, Sandia National Laboratories, the University of California at Berkeley, and the University of Tennessee. Our part of the project involves developing performance models for highly scalable algorithms and the development of latency tolerant iterative methods. During this project, we extended our performance models for the Multigrid method for solving large systems of linear equations and conducted experiments with highly scalable variants of conjugate gradient methods that avoid blocking synchronization. In addition, we worked with the other members of the project on alternative techniques for resilience and reproducibility.more » We also presented an alternative approach for reproducible dot-products in parallel computations that performs almost as well as the conventional approach by separating the order of computation from the details of the decomposition of vectors across the processes.« less

A scalable silicon photonic chip-scale optical switch for high performance computing systems.

PubMed

Yu, Runxiang; Cheung, Stanley; Li, Yuliang; Okamoto, Katsunari; Proietti, Roberto; Yin, Yawei; Yoo, S J B

2013-12-30

This paper discusses the architecture and provides performance studies of a silicon photonic chip-scale optical switch for scalable interconnect network in high performance computing systems. The proposed switch exploits optical wavelength parallelism and wavelength routing characteristics of an Arrayed Waveguide Grating Router (AWGR) to allow contention resolution in the wavelength domain. Simulation results from a cycle-accurate network simulator indicate that, even with only two transmitter/receiver pairs per node, the switch exhibits lower end-to-end latency and higher throughput at high (>90%) input loads compared with electronic switches. On the device integration level, we propose to integrate all the components (ring modulators, photodetectors and AWGR) on a CMOS-compatible silicon photonic platform to ensure a compact, energy efficient and cost-effective device. We successfully demonstrate proof-of-concept routing functions on an 8 × 8 prototype fabricated using foundry services provided by OpSIS-IME.

fastBMA: scalable network inference and transitive reduction.

PubMed

Hung, Ling-Hong; Shi, Kaiyuan; Wu, Migao; Young, William Chad; Raftery, Adrian E; Yeung, Ka Yee

2017-10-01

Inferring genetic networks from genome-wide expression data is extremely demanding computationally. We have developed fastBMA, a distributed, parallel, and scalable implementation of Bayesian model averaging (BMA) for this purpose. fastBMA also includes a computationally efficient module for eliminating redundant indirect edges in the network by mapping the transitive reduction to an easily solved shortest-path problem. We evaluated the performance of fastBMA on synthetic data and experimental genome-wide time series yeast and human datasets. When using a single CPU core, fastBMA is up to 100 times faster than the next fastest method, LASSO, with increased accuracy. It is a memory-efficient, parallel, and distributed application that scales to human genome-wide expression data. A 10 000-gene regulation network can be obtained in a matter of hours using a 32-core cloud cluster (2 nodes of 16 cores). fastBMA is a significant improvement over its predecessor ScanBMA. It is more accurate and orders of magnitude faster than other fast network inference methods such as the 1 based on LASSO. The improved scalability allows it to calculate networks from genome scale data in a reasonable time frame. The transitive reduction method can improve accuracy in denser networks. fastBMA is available as code (M.I.T. license) from GitHub (https://github.com/lhhunghimself/fastBMA), as part of the updated networkBMA Bioconductor package (https://www.bioconductor.org/packages/release/bioc/html/networkBMA.html) and as ready-to-deploy Docker images (https://hub.docker.com/r/biodepot/fastbma/). © The Authors 2017. Published by Oxford University Press.

Chebyshev polynomial filtered subspace iteration in the discontinuous Galerkin method for large-scale electronic structure calculations

DOE PAGES

Banerjee, Amartya S.; Lin, Lin; Hu, Wei; ...

2016-10-21

The Discontinuous Galerkin (DG) electronic structure method employs an adaptive local basis (ALB) set to solve the Kohn-Sham equations of density functional theory in a discontinuous Galerkin framework. The adaptive local basis is generated on-the-fly to capture the local material physics and can systematically attain chemical accuracy with only a few tens of degrees of freedom per atom. A central issue for large-scale calculations, however, is the computation of the electron density (and subsequently, ground state properties) from the discretized Hamiltonian in an efficient and scalable manner. We show in this work how Chebyshev polynomial filtered subspace iteration (CheFSI) canmore » be used to address this issue and push the envelope in large-scale materials simulations in a discontinuous Galerkin framework. We describe how the subspace filtering steps can be performed in an efficient and scalable manner using a two-dimensional parallelization scheme, thanks to the orthogonality of the DG basis set and block-sparse structure of the DG Hamiltonian matrix. The on-the-fly nature of the ALB functions requires additional care in carrying out the subspace iterations. We demonstrate the parallel scalability of the DG-CheFSI approach in calculations of large-scale twodimensional graphene sheets and bulk three-dimensional lithium-ion electrolyte systems. In conclusion, employing 55 296 computational cores, the time per self-consistent field iteration for a sample of the bulk 3D electrolyte containing 8586 atoms is 90 s, and the time for a graphene sheet containing 11 520 atoms is 75 s.« less

Hi-Corrector: a fast, scalable and memory-efficient package for normalizing large-scale Hi-C data.

PubMed

Li, Wenyuan; Gong, Ke; Li, Qingjiao; Alber, Frank; Zhou, Xianghong Jasmine

2015-03-15

Genome-wide proximity ligation assays, e.g. Hi-C and its variant TCC, have recently become important tools to study spatial genome organization. Removing biases from chromatin contact matrices generated by such techniques is a critical preprocessing step of subsequent analyses. The continuing decline of sequencing costs has led to an ever-improving resolution of the Hi-C data, resulting in very large matrices of chromatin contacts. Such large-size matrices, however, pose a great challenge on the memory usage and speed of its normalization. Therefore, there is an urgent need for fast and memory-efficient methods for normalization of Hi-C data. We developed Hi-Corrector, an easy-to-use, open source implementation of the Hi-C data normalization algorithm. Its salient features are (i) scalability-the software is capable of normalizing Hi-C data of any size in reasonable times; (ii) memory efficiency-the sequential version can run on any single computer with very limited memory, no matter how little; (iii) fast speed-the parallel version can run very fast on multiple computing nodes with limited local memory. The sequential version is implemented in ANSI C and can be easily compiled on any system; the parallel version is implemented in ANSI C with the MPI library (a standardized and portable parallel environment designed for solving large-scale scientific problems). The package is freely available at http://zhoulab.usc.edu/Hi-Corrector/. © The Author 2014. Published by Oxford University Press.

Development of the PARVMEC Code for Rapid Analysis of 3D MHD Equilibrium

NASA Astrophysics Data System (ADS)

Seal, Sudip; Hirshman, Steven; Cianciosa, Mark; Wingen, Andreas; Unterberg, Ezekiel; Wilcox, Robert; ORNL Collaboration

2015-11-01

The VMEC three-dimensional (3D) MHD equilibrium has been used extensively for designing stellarator experiments and analyzing experimental data in such strongly 3D systems. Recent applications of VMEC include 2D systems such as tokamaks (in particular, the D3D experiment), where application of very small (delB/B ~ 10-3) 3D resonant magnetic field perturbations render the underlying assumption of axisymmetry invalid. In order to facilitate the rapid analysis of such equilibria (for example, for reconstruction purposes), we have undertaken the task of parallelizing the VMEC code (PARVMEC) to produce a scalable and temporally rapidly convergent equilibrium code for use on parallel distributed memory platforms. The parallelization task naturally splits into three distinct parts 1) radial surfaces in the fixed-boundary part of the calculation; 2) two 2D angular meshes needed to compute the Green's function integrals over the plasma boundary for the free-boundary part of the code; and 3) block tridiagonal matrix needed to compute the full (3D) pre-conditioner near the final equilibrium state. Preliminary results show that scalability is achieved for tasks 1 and 3, with task 2 still nearing completion. The impact of this work on the rapid reconstruction of D3D plasmas using PARVMEC in the V3FIT code will be discussed. Work supported by U.S. DOE under Contract DE-AC05-00OR22725 with UT-Battelle, LLC.

Perspectives on the Future of CFD

NASA Technical Reports Server (NTRS)

Kwak, Dochan

2000-01-01

This viewgraph presentation gives an overview of the future of computational fluid dynamics (CFD), which in the past has pioneered the field of flow simulation. Over time CFD has progressed as computing power. Numerical methods have been advanced as CPU and memory capacity increases. Complex configurations are routinely computed now and direct numerical simulations (DNS) and large eddy simulations (LES) are used to study turbulence. As the computing resources changed to parallel and distributed platforms, computer science aspects such as scalability (algorithmic and implementation) and portability and transparent codings have advanced. Examples of potential future (or current) challenges include risk assessment, limitations of the heuristic model, and the development of CFD and information technology (IT) tools.

Modern gyrokinetic particle-in-cell simulation of fusion plasmas on top supercomputers

DOE PAGES

Wang, Bei; Ethier, Stephane; Tang, William; ...

2017-06-29

The Gyrokinetic Toroidal Code at Princeton (GTC-P) is a highly scalable and portable particle-in-cell (PIC) code. It solves the 5D Vlasov-Poisson equation featuring efficient utilization of modern parallel computer architectures at the petascale and beyond. Motivated by the goal of developing a modern code capable of dealing with the physics challenge of increasing problem size with sufficient resolution, new thread-level optimizations have been introduced as well as a key additional domain decomposition. GTC-P's multiple levels of parallelism, including inter-node 2D domain decomposition and particle decomposition, as well as intra-node shared memory partition and vectorization have enabled pushing the scalability ofmore » the PIC method to extreme computational scales. In this paper, we describe the methods developed to build a highly parallelized PIC code across a broad range of supercomputer designs. This particularly includes implementations on heterogeneous systems using NVIDIA GPU accelerators and Intel Xeon Phi (MIC) co-processors and performance comparisons with state-of-the-art homogeneous HPC systems such as Blue Gene/Q. New discovery science capabilities in the magnetic fusion energy application domain are enabled, including investigations of Ion-Temperature-Gradient (ITG) driven turbulence simulations with unprecedented spatial resolution and long temporal duration. Performance studies with realistic fusion experimental parameters are carried out on multiple supercomputing systems spanning a wide range of cache capacities, cache-sharing configurations, memory bandwidth, interconnects and network topologies. These performance comparisons using a realistic discovery-science-capable domain application code provide valuable insights on optimization techniques across one of the broadest sets of current high-end computing platforms worldwide.« less

Modern gyrokinetic particle-in-cell simulation of fusion plasmas on top supercomputers

DOE Office of Scientific and Technical Information (OSTI.GOV)

Wang, Bei; Ethier, Stephane; Tang, William

The Gyrokinetic Toroidal Code at Princeton (GTC-P) is a highly scalable and portable particle-in-cell (PIC) code. It solves the 5D Vlasov-Poisson equation featuring efficient utilization of modern parallel computer architectures at the petascale and beyond. Motivated by the goal of developing a modern code capable of dealing with the physics challenge of increasing problem size with sufficient resolution, new thread-level optimizations have been introduced as well as a key additional domain decomposition. GTC-P's multiple levels of parallelism, including inter-node 2D domain decomposition and particle decomposition, as well as intra-node shared memory partition and vectorization have enabled pushing the scalability ofmore » the PIC method to extreme computational scales. In this paper, we describe the methods developed to build a highly parallelized PIC code across a broad range of supercomputer designs. This particularly includes implementations on heterogeneous systems using NVIDIA GPU accelerators and Intel Xeon Phi (MIC) co-processors and performance comparisons with state-of-the-art homogeneous HPC systems such as Blue Gene/Q. New discovery science capabilities in the magnetic fusion energy application domain are enabled, including investigations of Ion-Temperature-Gradient (ITG) driven turbulence simulations with unprecedented spatial resolution and long temporal duration. Performance studies with realistic fusion experimental parameters are carried out on multiple supercomputing systems spanning a wide range of cache capacities, cache-sharing configurations, memory bandwidth, interconnects and network topologies. These performance comparisons using a realistic discovery-science-capable domain application code provide valuable insights on optimization techniques across one of the broadest sets of current high-end computing platforms worldwide.« less

Simple Linux Utility for Resource Management

DOE Office of Scientific and Technical Information (OSTI.GOV)

Jette, M.

2009-09-09

SLURM is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for large and small computer clusters. As a cluster resource manager, SLURM has three key functions. First, it allocates exclusive and/or non exclusive access to resources (compute nodes) to users for some duration of time so they can perform work. Second, it provides a framework for starting, executing, and monitoring work (normally a parallel job) on the set of allciated nodes. Finally, it arbitrates conflicting requests for resouces by managing a queue of pending work.

Integration experiences and performance studies of A COTS parallel archive systems

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chen, Hsing-bung; Scott, Cody; Grider, Bary

2010-01-01

Current and future Archive Storage Systems have been asked to (a) scale to very high bandwidths, (b) scale in metadata performance, (c) support policy-based hierarchical storage management capability, (d) scale in supporting changing needs of very large data sets, (e) support standard interface, and (f) utilize commercial-off-the-shelf(COTS) hardware. Parallel file systems have been asked to do the same thing but at one or more orders of magnitude faster in performance. Archive systems continue to move closer to file systems in their design due to the need for speed and bandwidth, especially metadata searching speeds such as more caching and lessmore » robust semantics. Currently the number of extreme highly scalable parallel archive solutions is very small especially those that will move a single large striped parallel disk file onto many tapes in parallel. We believe that a hybrid storage approach of using COTS components and innovative software technology can bring new capabilities into a production environment for the HPC community much faster than the approach of creating and maintaining a complete end-to-end unique parallel archive software solution. In this paper, we relay our experience of integrating a global parallel file system and a standard backup/archive product with a very small amount of additional code to provide a scalable, parallel archive. Our solution has a high degree of overlap with current parallel archive products including (a) doing parallel movement to/from tape for a single large parallel file, (b) hierarchical storage management, (c) ILM features, (d) high volume (non-single parallel file) archives for backup/archive/content management, and (e) leveraging all free file movement tools in Linux such as copy, move, ls, tar, etc. We have successfully applied our working COTS Parallel Archive System to the current world's first petaflop/s computing system, LANL's Roadrunner, and demonstrated its capability to address requirements of future archival storage systems.« less

Integration experiments and performance studies of a COTS parallel archive system

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chen, Hsing-bung; Scott, Cody; Grider, Gary

2010-06-16

Current and future Archive Storage Systems have been asked to (a) scale to very high bandwidths, (b) scale in metadata performance, (c) support policy-based hierarchical storage management capability, (d) scale in supporting changing needs of very large data sets, (e) support standard interface, and (f) utilize commercial-off-the-shelf (COTS) hardware. Parallel file systems have been asked to do the same thing but at one or more orders of magnitude faster in performance. Archive systems continue to move closer to file systems in their design due to the need for speed and bandwidth, especially metadata searching speeds such as more caching andmore » less robust semantics. Currently the number of extreme highly scalable parallel archive solutions is very small especially those that will move a single large striped parallel disk file onto many tapes in parallel. We believe that a hybrid storage approach of using COTS components and innovative software technology can bring new capabilities into a production environment for the HPC community much faster than the approach of creating and maintaining a complete end-to-end unique parallel archive software solution. In this paper, we relay our experience of integrating a global parallel file system and a standard backup/archive product with a very small amount of additional code to provide a scalable, parallel archive. Our solution has a high degree of overlap with current parallel archive products including (a) doing parallel movement to/from tape for a single large parallel file, (b) hierarchical storage management, (c) ILM features, (d) high volume (non-single parallel file) archives for backup/archive/content management, and (e) leveraging all free file movement tools in Linux such as copy, move, Is, tar, etc. We have successfully applied our working COTS Parallel Archive System to the current world's first petafiop/s computing system, LANL's Roadrunner machine, and demonstrated its capability to address requirements of future archival storage systems.« less

NWChem: A comprehensive and scalable open-source solution for large scale molecular simulations

NASA Astrophysics Data System (ADS)

Valiev, M.; Bylaska, E. J.; Govind, N.; Kowalski, K.; Straatsma, T. P.; Van Dam, H. J. J.; Wang, D.; Nieplocha, J.; Apra, E.; Windus, T. L.; de Jong, W. A.

2010-09-01

The latest release of NWChem delivers an open-source computational chemistry package with extensive capabilities for large scale simulations of chemical and biological systems. Utilizing a common computational framework, diverse theoretical descriptions can be used to provide the best solution for a given scientific problem. Scalable parallel implementations and modular software design enable efficient utilization of current computational architectures. This paper provides an overview of NWChem focusing primarily on the core theoretical modules provided by the code and their parallel performance. Program summaryProgram title: NWChem Catalogue identifier: AEGI_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEGI_v1_0.html Program obtainable from: CPC Program Library, Queen's University, Belfast, N. Ireland Licensing provisions: Open Source Educational Community License No. of lines in distributed program, including test data, etc.: 11 709 543 No. of bytes in distributed program, including test data, etc.: 680 696 106 Distribution format: tar.gz Programming language: Fortran 77, C Computer: all Linux based workstations and parallel supercomputers, Windows and Apple machines Operating system: Linux, OS X, Windows Has the code been vectorised or parallelized?: Code is parallelized Classification: 2.1, 2.2, 3, 7.3, 7.7, 16.1, 16.2, 16.3, 16.10, 16.13 Nature of problem: Large-scale atomistic simulations of chemical and biological systems require efficient and reliable methods for ground and excited solutions of many-electron Hamiltonian, analysis of the potential energy surface, and dynamics. Solution method: Ground and excited solutions of many-electron Hamiltonian are obtained utilizing density-functional theory, many-body perturbation approach, and coupled cluster expansion. These solutions or a combination thereof with classical descriptions are then used to analyze potential energy surface and perform dynamical simulations. Additional comments: Full documentation is provided in the distribution file. This includes an INSTALL file giving details of how to build the package. A set of test runs is provided in the examples directory. The distribution file for this program is over 90 Mbytes and therefore is not delivered directly when download or Email is requested. Instead a html file giving details of how the program can be obtained is sent. Running time: Running time depends on the size of the chemical system, complexity of the method, number of cpu's and the computational task. It ranges from several seconds for serial DFT energy calculations on a few atoms to several hours for parallel coupled cluster energy calculations on tens of atoms or ab-initio molecular dynamics simulation on hundreds of atoms.

PAGANI Toolkit: Parallel graph-theoretical analysis package for brain network big data.

PubMed

Du, Haixiao; Xia, Mingrui; Zhao, Kang; Liao, Xuhong; Yang, Huazhong; Wang, Yu; He, Yong

2018-05-01

The recent collection of unprecedented quantities of neuroimaging data with high spatial resolution has led to brain network big data. However, a toolkit for fast and scalable computational solutions is still lacking. Here, we developed the PArallel Graph-theoretical ANalysIs (PAGANI) Toolkit based on a hybrid central processing unit-graphics processing unit (CPU-GPU) framework with a graphical user interface to facilitate the mapping and characterization of high-resolution brain networks. Specifically, the toolkit provides flexible parameters for users to customize computations of graph metrics in brain network analyses. As an empirical example, the PAGANI Toolkit was applied to individual voxel-based brain networks with ∼200,000 nodes that were derived from a resting-state fMRI dataset of 624 healthy young adults from the Human Connectome Project. Using a personal computer, this toolbox completed all computations in ∼27 h for one subject, which is markedly less than the 118 h required with a single-thread implementation. The voxel-based functional brain networks exhibited prominent small-world characteristics and densely connected hubs, which were mainly located in the medial and lateral fronto-parietal cortices. Moreover, the female group had significantly higher modularity and nodal betweenness centrality mainly in the medial/lateral fronto-parietal and occipital cortices than the male group. Significant correlations between the intelligence quotient and nodal metrics were also observed in several frontal regions. Collectively, the PAGANI Toolkit shows high computational performance and good scalability for analyzing connectome big data and provides a friendly interface without the complicated configuration of computing environments, thereby facilitating high-resolution connectomics research in health and disease. © 2018 Wiley Periodicals, Inc.

NETRA: A parallel architecture for integrated vision systems. 1: Architecture and organization

NASA Technical Reports Server (NTRS)

Choudhary, Alok N.; Patel, Janak H.; Ahuja, Narendra

1989-01-01

Computer vision is regarded as one of the most complex and computationally intensive problems. An integrated vision system (IVS) is considered to be a system that uses vision algorithms from all levels of processing for a high level application (such as object recognition). A model of computation is presented for parallel processing for an IVS. Using the model, desired features and capabilities of a parallel architecture suitable for IVSs are derived. Then a multiprocessor architecture (called NETRA) is presented. This architecture is highly flexible without the use of complex interconnection schemes. The topology of NETRA is recursively defined and hence is easily scalable from small to large systems. Homogeneity of NETRA permits fault tolerance and graceful degradation under faults. It is a recursively defined tree-type hierarchical architecture where each of the leaf nodes consists of a cluster of processors connected with a programmable crossbar with selective broadcast capability to provide for desired flexibility. A qualitative evaluation of NETRA is presented. Then general schemes are described to map parallel algorithms onto NETRA. Algorithms are classified according to their communication requirements for parallel processing. An extensive analysis of inter-cluster communication strategies in NETRA is presented, and parameters affecting performance of parallel algorithms when mapped on NETRA are discussed. Finally, a methodology to evaluate performance of algorithms on NETRA is described.

Design of a dataway processor for a parallel image signal processing system

NASA Astrophysics Data System (ADS)

Nomura, Mitsuru; Fujii, Tetsuro; Ono, Sadayasu

1995-04-01

Recently, demands for high-speed signal processing have been increasing especially in the field of image data compression, computer graphics, and medical imaging. To achieve sufficient power for real-time image processing, we have been developing parallel signal-processing systems. This paper describes a communication processor called 'dataway processor' designed for a new scalable parallel signal-processing system. The processor has six high-speed communication links (Dataways), a data-packet routing controller, a RISC CORE, and a DMA controller. Each communication link operates at 8-bit parallel in a full duplex mode at 50 MHz. Moreover, data routing, DMA, and CORE operations are processed in parallel. Therefore, sufficient throughput is available for high-speed digital video signals. The processor is designed in a top- down fashion using a CAD system called 'PARTHENON.' The hardware is fabricated using 0.5-micrometers CMOS technology, and its hardware is about 200 K gates.

Parallel processor for real-time structural control

NASA Astrophysics Data System (ADS)

Tise, Bert L.

1993-07-01

A parallel processor that is optimized for real-time linear control has been developed. This modular system consists of A/D modules, D/A modules, and floating-point processor modules. The scalable processor uses up to 1,000 Motorola DSP96002 floating-point processors for a peak computational rate of 60 GFLOPS. Sampling rates up to 625 kHz are supported by this analog-in to analog-out controller. The high processing rate and parallel architecture make this processor suitable for computing state-space equations and other multiply/accumulate-intensive digital filters. Processor features include 14-bit conversion devices, low input-to-output latency, 240 Mbyte/s synchronous backplane bus, low-skew clock distribution circuit, VME connection to host computer, parallelizing code generator, and look- up-tables for actuator linearization. This processor was designed primarily for experiments in structural control. The A/D modules sample sensors mounted on the structure and the floating- point processor modules compute the outputs using the programmed control equations. The outputs are sent through the D/A module to the power amps used to drive the structure's actuators. The host computer is a Sun workstation. An OpenWindows-based control panel is provided to facilitate data transfer to and from the processor, as well as to control the operating mode of the processor. A diagnostic mode is provided to allow stimulation of the structure and acquisition of the structural response via sensor inputs.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Barrett, Brian; Brightwell, Ronald B.; Grant, Ryan

This report presents a specification for the Portals 4 networ k programming interface. Portals 4 is intended to allow scalable, high-performance network communication betwee n nodes of a parallel computing system. Portals 4 is well suited to massively parallel processing and embedded syste ms. Portals 4 represents an adaption of the data movement layer developed for massively parallel processing platfor ms, such as the 4500-node Intel TeraFLOPS machine. Sandia's Cplant cluster project motivated the development of Version 3.0, which was later extended to Version 3.3 as part of the Cray Red Storm machine and XT line. Version 4 is tarmore » geted to the next generation of machines employing advanced network interface architectures that support enh anced offload capabilities.« less

The Portals 4.0 network programming interface.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Barrett, Brian W.; Brightwell, Ronald Brian; Pedretti, Kevin

2012-11-01

This report presents a specification for the Portals 4.0 network programming interface. Portals 4.0 is intended to allow scalable, high-performance network communication between nodes of a parallel computing system. Portals 4.0 is well suited to massively parallel processing and embedded systems. Portals 4.0 represents an adaption of the data movement layer developed for massively parallel processing platforms, such as the 4500-node Intel TeraFLOPS machine. Sandias Cplant cluster project motivated the development of Version 3.0, which was later extended to Version 3.3 as part of the Cray Red Storm machine and XT line. Version 4.0 is targeted to the next generationmore » of machines employing advanced network interface architectures that support enhanced offload capabilities.« less

SBML-PET-MPI: a parallel parameter estimation tool for Systems Biology Markup Language based models.

PubMed

Zi, Zhike

2011-04-01

Parameter estimation is crucial for the modeling and dynamic analysis of biological systems. However, implementing parameter estimation is time consuming and computationally demanding. Here, we introduced a parallel parameter estimation tool for Systems Biology Markup Language (SBML)-based models (SBML-PET-MPI). SBML-PET-MPI allows the user to perform parameter estimation and parameter uncertainty analysis by collectively fitting multiple experimental datasets. The tool is developed and parallelized using the message passing interface (MPI) protocol, which provides good scalability with the number of processors. SBML-PET-MPI is freely available for non-commercial use at http://www.bioss.uni-freiburg.de/cms/sbml-pet-mpi.html or http://sites.google.com/site/sbmlpetmpi/.

Shift-and-invert parallel spectral transformation eigensolver: Massively parallel performance for density-functional based tight-binding

DOE PAGES

Zhang, Hong; Zapol, Peter; Dixon, David A.; ...

2015-11-17

The Shift-and-invert parallel spectral transformations (SIPs), a computational approach to solve sparse eigenvalue problems, is developed for massively parallel architectures with exceptional parallel scalability and robustness. The capabilities of SIPs are demonstrated by diagonalization of density-functional based tight-binding (DFTB) Hamiltonian and overlap matrices for single-wall metallic carbon nanotubes, diamond nanowires, and bulk diamond crystals. The largest (smallest) example studied is a 128,000 (2000) atom nanotube for which ~330,000 (~5600) eigenvalues and eigenfunctions are obtained in ~190 (~5) seconds when parallelized over 266,144 (16,384) Blue Gene/Q cores. Weak scaling and strong scaling of SIPs are analyzed and the performance of SIPsmore » is compared with other novel methods. Different matrix ordering methods are investigated to reduce the cost of the factorization step, which dominates the time-to-solution at the strong scaling limit. As a result, a parallel implementation of assembling the density matrix from the distributed eigenvectors is demonstrated.« less

Shift-and-invert parallel spectral transformation eigensolver: Massively parallel performance for density-functional based tight-binding

DOE Office of Scientific and Technical Information (OSTI.GOV)

Zhang, Hong; Zapol, Peter; Dixon, David A.

The Shift-and-invert parallel spectral transformations (SIPs), a computational approach to solve sparse eigenvalue problems, is developed for massively parallel architectures with exceptional parallel scalability and robustness. The capabilities of SIPs are demonstrated by diagonalization of density-functional based tight-binding (DFTB) Hamiltonian and overlap matrices for single-wall metallic carbon nanotubes, diamond nanowires, and bulk diamond crystals. The largest (smallest) example studied is a 128,000 (2000) atom nanotube for which ~330,000 (~5600) eigenvalues and eigenfunctions are obtained in ~190 (~5) seconds when parallelized over 266,144 (16,384) Blue Gene/Q cores. Weak scaling and strong scaling of SIPs are analyzed and the performance of SIPsmore » is compared with other novel methods. Different matrix ordering methods are investigated to reduce the cost of the factorization step, which dominates the time-to-solution at the strong scaling limit. As a result, a parallel implementation of assembling the density matrix from the distributed eigenvectors is demonstrated.« less

Implementation of BT, SP, LU, and FT of NAS Parallel Benchmarks in Java

NASA Technical Reports Server (NTRS)

Schultz, Matthew; Frumkin, Michael; Jin, Hao-Qiang; Yan, Jerry

2000-01-01

A number of Java features make it an attractive but a debatable choice for High Performance Computing. We have implemented benchmarks working on single structured grid BT,SP,LU and FT in Java. The performance and scalability of the Java code shows that a significant improvement in Java compiler technology and in Java thread implementation are necessary for Java to compete with Fortran in HPC applications.

Multi-mode sensor processing on a dynamically reconfigurable massively parallel processor array

NASA Astrophysics Data System (ADS)

Chen, Paul; Butts, Mike; Budlong, Brad; Wasson, Paul

2008-04-01

This paper introduces a novel computing architecture that can be reconfigured in real time to adapt on demand to multi-mode sensor platforms' dynamic computational and functional requirements. This 1 teraOPS reconfigurable Massively Parallel Processor Array (MPPA) has 336 32-bit processors. The programmable 32-bit communication fabric provides streamlined inter-processor connections with deterministically high performance. Software programmability, scalability, ease of use, and fast reconfiguration time (ranging from microseconds to milliseconds) are the most significant advantages over FPGAs and DSPs. This paper introduces the MPPA architecture, its programming model, and methods of reconfigurability. An MPPA platform for reconfigurable computing is based on a structural object programming model. Objects are software programs running concurrently on hundreds of 32-bit RISC processors and memories. They exchange data and control through a network of self-synchronizing channels. A common application design pattern on this platform, called a work farm, is a parallel set of worker objects, with one input and one output stream. Statically configured work farms with homogeneous and heterogeneous sets of workers have been used in video compression and decompression, network processing, and graphics applications.

SNAVA-A real-time multi-FPGA multi-model spiking neural network simulation architecture.

PubMed

Sripad, Athul; Sanchez, Giovanny; Zapata, Mireya; Pirrone, Vito; Dorta, Taho; Cambria, Salvatore; Marti, Albert; Krishnamourthy, Karthikeyan; Madrenas, Jordi

2018-01-01

Spiking Neural Networks (SNN) for Versatile Applications (SNAVA) simulation platform is a scalable and programmable parallel architecture that supports real-time, large-scale, multi-model SNN computation. This parallel architecture is implemented in modern Field-Programmable Gate Arrays (FPGAs) devices to provide high performance execution and flexibility to support large-scale SNN models. Flexibility is defined in terms of programmability, which allows easy synapse and neuron implementation. This has been achieved by using a special-purpose Processing Elements (PEs) for computing SNNs, and analyzing and customizing the instruction set according to the processing needs to achieve maximum performance with minimum resources. The parallel architecture is interfaced with customized Graphical User Interfaces (GUIs) to configure the SNN's connectivity, to compile the neuron-synapse model and to monitor SNN's activity. Our contribution intends to provide a tool that allows to prototype SNNs faster than on CPU/GPU architectures but significantly cheaper than fabricating a customized neuromorphic chip. This could be potentially valuable to the computational neuroscience and neuromorphic engineering communities. Copyright © 2017 Elsevier Ltd. All rights reserved.

Integrating the Apache Big Data Stack with HPC for Big Data

NASA Astrophysics Data System (ADS)

Fox, G. C.; Qiu, J.; Jha, S.

2014-12-01

There is perhaps a broad consensus as to important issues in practical parallel computing as applied to large scale simulations; this is reflected in supercomputer architectures, algorithms, libraries, languages, compilers and best practice for application development. However, the same is not so true for data intensive computing, even though commercially clouds devote much more resources to data analytics than supercomputers devote to simulations. We look at a sample of over 50 big data applications to identify characteristics of data intensive applications and to deduce needed runtime and architectures. We suggest a big data version of the famous Berkeley dwarfs and NAS parallel benchmarks and use these to identify a few key classes of hardware/software architectures. Our analysis builds on combining HPC and ABDS the Apache big data software stack that is well used in modern cloud computing. Initial results on clouds and HPC systems are encouraging. We propose the development of SPIDAL - Scalable Parallel Interoperable Data Analytics Library -- built on system aand data abstractions suggested by the HPC-ABDS architecture. We discuss how it can be used in several application areas including Polar Science.

Message Passing and Shared Address Space Parallelism on an SMP Cluster

NASA Technical Reports Server (NTRS)

Shan, Hongzhang; Singh, Jaswinder P.; Oliker, Leonid; Biswas, Rupak; Biegel, Bryan (Technical Monitor)

2002-01-01

Currently, message passing (MP) and shared address space (SAS) are the two leading parallel programming paradigms. MP has been standardized with MPI, and is the more common and mature approach; however, code development can be extremely difficult, especially for irregularly structured computations. SAS offers substantial ease of programming, but may suffer from performance limitations due to poor spatial locality and high protocol overhead. In this paper, we compare the performance of and the programming effort required for six applications under both programming models on a 32-processor PC-SMP cluster, a platform that is becoming increasingly attractive for high-end scientific computing. Our application suite consists of codes that typically do not exhibit scalable performance under shared-memory programming due to their high communication-to-computation ratios and/or complex communication patterns. Results indicate that SAS can achieve about half the parallel efficiency of MPI for most of our applications, while being competitive for the others. A hybrid MPI+SAS strategy shows only a small performance advantage over pure MPI in some cases. Finally, improved implementations of two MPI collective operations on PC-SMP clusters are presented.

A scalable nonlinear fluid-structure interaction solver based on a Schwarz preconditioner with isogeometric unstructured coarse spaces in 3D

NASA Astrophysics Data System (ADS)

Kong, Fande; Cai, Xiao-Chuan

2017-07-01

Nonlinear fluid-structure interaction (FSI) problems on unstructured meshes in 3D appear in many applications in science and engineering, such as vibration analysis of aircrafts and patient-specific diagnosis of cardiovascular diseases. In this work, we develop a highly scalable, parallel algorithmic and software framework for FSI problems consisting of a nonlinear fluid system and a nonlinear solid system, that are coupled monolithically. The FSI system is discretized by a stabilized finite element method in space and a fully implicit backward difference scheme in time. To solve the large, sparse system of nonlinear algebraic equations at each time step, we propose an inexact Newton-Krylov method together with a multilevel, smoothed Schwarz preconditioner with isogeometric coarse meshes generated by a geometry preserving coarsening algorithm. Here "geometry" includes the boundary of the computational domain and the wet interface between the fluid and the solid. We show numerically that the proposed algorithm and implementation are highly scalable in terms of the number of linear and nonlinear iterations and the total compute time on a supercomputer with more than 10,000 processor cores for several problems with hundreds of millions of unknowns.

A scalable nonlinear fluid–structure interaction solver based on a Schwarz preconditioner with isogeometric unstructured coarse spaces in 3D

DOE PAGES

Kong, Fande; Cai, Xiao-Chuan

2017-03-24

Nonlinear fluid-structure interaction (FSI) problems on unstructured meshes in 3D appear many applications in science and engineering, such as vibration analysis of aircrafts and patient-specific diagnosis of cardiovascular diseases. In this work, we develop a highly scalable, parallel algorithmic and software framework for FSI problems consisting of a nonlinear fluid system and a nonlinear solid system, that are coupled monolithically. The FSI system is discretized by a stabilized finite element method in space and a fully implicit backward difference scheme in time. To solve the large, sparse system of nonlinear algebraic equations at each time step, we propose an inexactmore » Newton-Krylov method together with a multilevel, smoothed Schwarz preconditioner with isogeometric coarse meshes generated by a geometry preserving coarsening algorithm. Here ''geometry'' includes the boundary of the computational domain and the wet interface between the fluid and the solid. We show numerically that the proposed algorithm and implementation are highly scalable in terms of the number of linear and nonlinear iterations and the total compute time on a supercomputer with more than 10,000 processor cores for several problems with hundreds of millions of unknowns.« less

Parallel algorithm of VLBI software correlator under multiprocessor environment

NASA Astrophysics Data System (ADS)

Zheng, Weimin; Zhang, Dong

2007-11-01

The correlator is the key signal processing equipment of a Very Lone Baseline Interferometry (VLBI) synthetic aperture telescope. It receives the mass data collected by the VLBI observatories and produces the visibility function of the target, which can be used to spacecraft position, baseline length measurement, synthesis imaging, and other scientific applications. VLBI data correlation is a task of data intensive and computation intensive. This paper presents the algorithms of two parallel software correlators under multiprocessor environments. A near real-time correlator for spacecraft tracking adopts the pipelining and thread-parallel technology, and runs on the SMP (Symmetric Multiple Processor) servers. Another high speed prototype correlator using the mixed Pthreads and MPI (Massage Passing Interface) parallel algorithm is realized on a small Beowulf cluster platform. Both correlators have the characteristic of flexible structure, scalability, and with 10-station data correlating abilities.

Massively parallel and linear-scaling algorithm for second-order Møller-Plesset perturbation theory applied to the study of supramolecular wires

NASA Astrophysics Data System (ADS)

Kjærgaard, Thomas; Baudin, Pablo; Bykov, Dmytro; Eriksen, Janus Juul; Ettenhuber, Patrick; Kristensen, Kasper; Larkin, Jeff; Liakh, Dmitry; Pawłowski, Filip; Vose, Aaron; Wang, Yang Min; Jørgensen, Poul

2017-03-01

We present a scalable cross-platform hybrid MPI/OpenMP/OpenACC implementation of the Divide-Expand-Consolidate (DEC) formalism with portable performance on heterogeneous HPC architectures. The Divide-Expand-Consolidate formalism is designed to reduce the steep computational scaling of conventional many-body methods employed in electronic structure theory to linear scaling, while providing a simple mechanism for controlling the error introduced by this approximation. Our massively parallel implementation of this general scheme has three levels of parallelism, being a hybrid of the loosely coupled task-based parallelization approach and the conventional MPI +X programming model, where X is either OpenMP or OpenACC. We demonstrate strong and weak scalability of this implementation on heterogeneous HPC systems, namely on the GPU-based Cray XK7 Titan supercomputer at the Oak Ridge National Laboratory. Using the "resolution of the identity second-order Møller-Plesset perturbation theory" (RI-MP2) as the physical model for simulating correlated electron motion, the linear-scaling DEC implementation is applied to 1-aza-adamantane-trione (AAT) supramolecular wires containing up to 40 monomers (2440 atoms, 6800 correlated electrons, 24 440 basis functions and 91 280 auxiliary functions). This represents the largest molecular system treated at the MP2 level of theory, demonstrating an efficient removal of the scaling wall pertinent to conventional quantum many-body methods.

Reducing Communication in Algebraic Multigrid Using Additive Variants

DOE Office of Scientific and Technical Information (OSTI.GOV)

Vassilevski, Panayot S.; Yang, Ulrike Meier

Algebraic multigrid (AMG) has proven to be an effective scalable solver on many high performance computers. However, its increasing communication complexity on coarser levels has shown to seriously impact its performance on computers with high communication cost. Moreover, additive AMG variants provide not only increased parallelism as well as decreased numbers of messages per cycle but also generally exhibit slower convergence. Here we present various new additive variants with convergence rates that are significantly improved compared to the classical additive algebraic multigrid method and investigate their potential for decreased communication, and improved communication-computation overlap, features that are essential for goodmore » performance on future exascale architectures.« less

Reducing Communication in Algebraic Multigrid Using Additive Variants

DOE PAGES

Vassilevski, Panayot S.; Yang, Ulrike Meier

2014-02-12

Algebraic multigrid (AMG) has proven to be an effective scalable solver on many high performance computers. However, its increasing communication complexity on coarser levels has shown to seriously impact its performance on computers with high communication cost. Moreover, additive AMG variants provide not only increased parallelism as well as decreased numbers of messages per cycle but also generally exhibit slower convergence. Here we present various new additive variants with convergence rates that are significantly improved compared to the classical additive algebraic multigrid method and investigate their potential for decreased communication, and improved communication-computation overlap, features that are essential for goodmore » performance on future exascale architectures.« less

Large Scale Flutter Data for Design of Rotating Blades Using Navier-Stokes Equations

NASA Technical Reports Server (NTRS)

Guruswamy, Guru P.

2012-01-01

A procedure to compute flutter boundaries of rotating blades is presented; a) Navier-Stokes equations. b) Frequency domain method compatible with industry practice. Procedure is initially validated: a) Unsteady loads with flapping wing experiment. b) Flutter boundary with fixed wing experiment. Large scale flutter computation is demonstrated for rotating blade: a) Single job submission script. b) Flutter boundary in 24 hour wall clock time with 100 cores. c) Linearly scalable with number of cores. Tested with 1000 cores that produced data in 25 hrs for 10 flutter boundaries. Further wall-clock speed-up is possible by performing parallel computations within each case.

Lattice dynamics calculations based on density-functional perturbation theory in real space

NASA Astrophysics Data System (ADS)

Shang, Honghui; Carbogno, Christian; Rinke, Patrick; Scheffler, Matthias

2017-06-01

A real-space formalism for density-functional perturbation theory (DFPT) is derived and applied for the computation of harmonic vibrational properties in molecules and solids. The practical implementation using numeric atom-centered orbitals as basis functions is demonstrated exemplarily for the all-electron Fritz Haber Institute ab initio molecular simulations (FHI-aims) package. The convergence of the calculations with respect to numerical parameters is carefully investigated and a systematic comparison with finite-difference approaches is performed both for finite (molecules) and extended (periodic) systems. Finally, the scaling tests and scalability tests on massively parallel computer systems demonstrate the computational efficiency.

Smart integrated microsystems: the energy efficiency challenge (Conference Presentation) (Plenary Presentation)

NASA Astrophysics Data System (ADS)

Benini, Luca

2017-06-01

The "internet of everything" envisions trillions of connected objects loaded with high-bandwidth sensors requiring massive amounts of local signal processing, fusion, pattern extraction and classification. From the computational viewpoint, the challenge is formidable and can be addressed only by pushing computing fabrics toward massive parallelism and brain-like energy efficiency levels. CMOS technology can still take us a long way toward this goal, but technology scaling is losing steam. Energy efficiency improvement will increasingly hinge on architecture, circuits, design techniques such as heterogeneous 3D integration, mixed-signal preprocessing, event-based approximate computing and non-Von-Neumann architectures for scalable acceleration.

Application of a hybrid MPI/OpenMP approach for parallel groundwater model calibration using multi-core computers

DOE Office of Scientific and Technical Information (OSTI.GOV)

Tang, Guoping; D'Azevedo, Ed F; Zhang, Fan

2010-01-01

Calibration of groundwater models involves hundreds to thousands of forward solutions, each of which may solve many transient coupled nonlinear partial differential equations, resulting in a computationally intensive problem. We describe a hybrid MPI/OpenMP approach to exploit two levels of parallelisms in software and hardware to reduce calibration time on multi-core computers. HydroGeoChem 5.0 (HGC5) is parallelized using OpenMP for direct solutions for a reactive transport model application, and a field-scale coupled flow and transport model application. In the reactive transport model, a single parallelizable loop is identified to account for over 97% of the total computational time using GPROF.more » Addition of a few lines of OpenMP compiler directives to the loop yields a speedup of about 10 on a 16-core compute node. For the field-scale model, parallelizable loops in 14 of 174 HGC5 subroutines that require 99% of the execution time are identified. As these loops are parallelized incrementally, the scalability is found to be limited by a loop where Cray PAT detects over 90% cache missing rates. With this loop rewritten, similar speedup as the first application is achieved. The OpenMP-parallelized code can be run efficiently on multiple workstations in a network or multiple compute nodes on a cluster as slaves using parallel PEST to speedup model calibration. To run calibration on clusters as a single task, the Levenberg Marquardt algorithm is added to HGC5 with the Jacobian calculation and lambda search parallelized using MPI. With this hybrid approach, 100 200 compute cores are used to reduce the calibration time from weeks to a few hours for these two applications. This approach is applicable to most of the existing groundwater model codes for many applications.« less

Towards scalable Byzantine fault-tolerant replication

NASA Astrophysics Data System (ADS)

Zbierski, Maciej

2017-08-01

Byzantine fault-tolerant (BFT) replication is a powerful technique, enabling distributed systems to remain available and correct even in the presence of arbitrary faults. Unfortunately, existing BFT replication protocols are mostly load-unscalable, i.e. they fail to respond with adequate performance increase whenever new computational resources are introduced into the system. This article proposes a universal architecture facilitating the creation of load-scalable distributed services based on BFT replication. The suggested approach exploits parallel request processing to fully utilize the available resources, and uses a load balancer module to dynamically adapt to the properties of the observed client workload. The article additionally provides a discussion on selected deployment scenarios, and explains how the proposed architecture could be used to increase the dependability of contemporary large-scale distributed systems.

Tinker-HP: a massively parallel molecular dynamics package for multiscale simulations of large complex systems with advanced point dipole polarizable force fields.

PubMed

Lagardère, Louis; Jolly, Luc-Henri; Lipparini, Filippo; Aviat, Félix; Stamm, Benjamin; Jing, Zhifeng F; Harger, Matthew; Torabifard, Hedieh; Cisneros, G Andrés; Schnieders, Michael J; Gresh, Nohad; Maday, Yvon; Ren, Pengyu Y; Ponder, Jay W; Piquemal, Jean-Philip

2018-01-28

We present Tinker-HP, a massively MPI parallel package dedicated to classical molecular dynamics (MD) and to multiscale simulations, using advanced polarizable force fields (PFF) encompassing distributed multipoles electrostatics. Tinker-HP is an evolution of the popular Tinker package code that conserves its simplicity of use and its reference double precision implementation for CPUs. Grounded on interdisciplinary efforts with applied mathematics, Tinker-HP allows for long polarizable MD simulations on large systems up to millions of atoms. We detail in the paper the newly developed extension of massively parallel 3D spatial decomposition to point dipole polarizable models as well as their coupling to efficient Krylov iterative and non-iterative polarization solvers. The design of the code allows the use of various computer systems ranging from laboratory workstations to modern petascale supercomputers with thousands of cores. Tinker-HP proposes therefore the first high-performance scalable CPU computing environment for the development of next generation point dipole PFFs and for production simulations. Strategies linking Tinker-HP to Quantum Mechanics (QM) in the framework of multiscale polarizable self-consistent QM/MD simulations are also provided. The possibilities, performances and scalability of the software are demonstrated via benchmarks calculations using the polarizable AMOEBA force field on systems ranging from large water boxes of increasing size and ionic liquids to (very) large biosystems encompassing several proteins as well as the complete satellite tobacco mosaic virus and ribosome structures. For small systems, Tinker-HP appears to be competitive with the Tinker-OpenMM GPU implementation of Tinker. As the system size grows, Tinker-HP remains operational thanks to its access to distributed memory and takes advantage of its new algorithmic enabling for stable long timescale polarizable simulations. Overall, a several thousand-fold acceleration over a single-core computation is observed for the largest systems. The extension of the present CPU implementation of Tinker-HP to other computational platforms is discussed.

al3c: high-performance software for parameter inference using Approximate Bayesian Computation.

PubMed

Stram, Alexander H; Marjoram, Paul; Chen, Gary K

2015-11-01

The development of Approximate Bayesian Computation (ABC) algorithms for parameter inference which are both computationally efficient and scalable in parallel computing environments is an important area of research. Monte Carlo rejection sampling, a fundamental component of ABC algorithms, is trivial to distribute over multiple processors but is inherently inefficient. While development of algorithms such as ABC Sequential Monte Carlo (ABC-SMC) help address the inherent inefficiencies of rejection sampling, such approaches are not as easily scaled on multiple processors. As a result, current Bayesian inference software offerings that use ABC-SMC lack the ability to scale in parallel computing environments. We present al3c, a C++ framework for implementing ABC-SMC in parallel. By requiring only that users define essential functions such as the simulation model and prior distribution function, al3c abstracts the user from both the complexities of parallel programming and the details of the ABC-SMC algorithm. By using the al3c framework, the user is able to scale the ABC-SMC algorithm in parallel computing environments for his or her specific application, with minimal programming overhead. al3c is offered as a static binary for Linux and OS-X computing environments. The user completes an XML configuration file and C++ plug-in template for the specific application, which are used by al3c to obtain the desired results. Users can download the static binaries, source code, reference documentation and examples (including those in this article) by visiting https://github.com/ahstram/al3c. astram@usc.edu Supplementary data are available at Bioinformatics online. © The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Parallel auto-correlative statistics with VTK.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Pebay, Philippe Pierre; Bennett, Janine Camille

2013-08-01

This report summarizes existing statistical engines in VTK and presents both the serial and parallel auto-correlative statistics engines. It is a sequel to [PT08, BPRT09b, PT09, BPT09, PT10] which studied the parallel descriptive, correlative, multi-correlative, principal component analysis, contingency, k-means, and order statistics engines. The ease of use of the new parallel auto-correlative statistics engine is illustrated by the means of C++ code snippets and algorithm verification is provided. This report justifies the design of the statistics engines with parallel scalability in mind, and provides scalability and speed-up analysis results for the autocorrelative statistics engine.

PETSc Users Manual Revision 3.3

DOE Office of Scientific and Technical Information (OSTI.GOV)

Balay, S.; Brown, J.; Buschelman, K.

This manual describes the use of PETSc for the numerical solution of partial differential equations and related problems on high-performance computers. The Portable, Extensible Toolkit for Scientific Computation (PETSc) is a suite of data structures and routines that provide the building blocks for the implementation of large-scale application codes on parallel (and serial) computers. PETSc uses the MPI standard for all message-passing communication. PETSc includes an expanding suite of parallel linear, nonlinear equation solvers and time integrators that may be used in application codes written in Fortran, C, C++, Python, and MATLAB (sequential). PETSc provides many of the mechanisms neededmore » within parallel application codes, such as parallel matrix and vector assembly routines. The library is organized hierarchically, enabling users to employ the level of abstraction that is most appropriate for a particular problem. By using techniques of object-oriented programming, PETSc provides enormous flexibility for users. PETSc is a sophisticated set of software tools; as such, for some users it initially has a much steeper learning curve than a simple subroutine library. In particular, for individuals without some computer science background, experience programming in C, C++ or Fortran and experience using a debugger such as gdb or dbx, it may require a significant amount of time to take full advantage of the features that enable efficient software use. However, the power of the PETSc design and the algorithms it incorporates may make the efficient implementation of many application codes simpler than “rolling them” yourself; For many tasks a package such as MATLAB is often the best tool; PETSc is not intended for the classes of problems for which effective MATLAB code can be written. PETSc also has a MATLAB interface, so portions of your code can be written in MATLAB to “try out” the PETSc solvers. The resulting code will not be scalable however because currently MATLAB is inherently not scalable; and PETSc should not be used to attempt to provide a “parallel linear solver” in an otherwise sequential code. Certainly all parts of a previously sequential code need not be parallelized but the matrix generation portion must be parallelized to expect any kind of reasonable performance. Do not expect to generate your matrix sequentially and then “use PETSc” to solve the linear system in parallel. Since PETSc is under continued development, small changes in usage and calling sequences of routines will occur. PETSc is supported; see the web site http://www.mcs.anl.gov/petsc for information on contacting support. A http://www.mcs.anl.gov/petsc/publications may be found a list of publications and web sites that feature work involving PETSc. We welcome any reports of corrections for this document.« less

PETSc Users Manual Revision 3.4

DOE Office of Scientific and Technical Information (OSTI.GOV)

Balay, S.; Brown, J.; Buschelman, K.

This manual describes the use of PETSc for the numerical solution of partial differential equations and related problems on high-performance computers. The Portable, Extensible Toolkit for Scientific Computation (PETSc) is a suite of data structures and routines that provide the building blocks for the implementation of large-scale application codes on parallel (and serial) computers. PETSc uses the MPI standard for all message-passing communication. PETSc includes an expanding suite of parallel linear, nonlinear equation solvers and time integrators that may be used in application codes written in Fortran, C, C++, Python, and MATLAB (sequential). PETSc provides many of the mechanisms neededmore » within parallel application codes, such as parallel matrix and vector assembly routines. The library is organized hierarchically, enabling users to employ the level of abstraction that is most appropriate for a particular problem. By using techniques of object-oriented programming, PETSc provides enormous flexibility for users. PETSc is a sophisticated set of software tools; as such, for some users it initially has a much steeper learning curve than a simple subroutine library. In particular, for individuals without some computer science background, experience programming in C, C++ or Fortran and experience using a debugger such as gdb or dbx, it may require a significant amount of time to take full advantage of the features that enable efficient software use. However, the power of the PETSc design and the algorithms it incorporates may make the efficient implementation of many application codes simpler than “rolling them” yourself; For many tasks a package such as MATLAB is often the best tool; PETSc is not intended for the classes of problems for which effective MATLAB code can be written. PETSc also has a MATLAB interface, so portions of your code can be written in MATLAB to “try out” the PETSc solvers. The resulting code will not be scalable however because currently MATLAB is inherently not scalable; and PETSc should not be used to attempt to provide a “parallel linear solver” in an otherwise sequential code. Certainly all parts of a previously sequential code need not be parallelized but the matrix generation portion must be parallelized to expect any kind of reasonable performance. Do not expect to generate your matrix sequentially and then “use PETSc” to solve the linear system in parallel. Since PETSc is under continued development, small changes in usage and calling sequences of routines will occur. PETSc is supported; see the web site http://www.mcs.anl.gov/petsc for information on contacting support. A http://www.mcs.anl.gov/petsc/publications may be found a list of publications and web sites that feature work involving PETSc. We welcome any reports of corrections for this document.« less

PETSc Users Manual Revision 3.5

DOE Office of Scientific and Technical Information (OSTI.GOV)

Balay, S.; Abhyankar, S.; Adams, M.

This manual describes the use of PETSc for the numerical solution of partial differential equations and related problems on high-performance computers. The Portable, Extensible Toolkit for Scientific Computation (PETSc) is a suite of data structures and routines that provide the building blocks for the implementation of large-scale application codes on parallel (and serial) computers. PETSc uses the MPI standard for all message-passing communication. PETSc includes an expanding suite of parallel linear, nonlinear equation solvers and time integrators that may be used in application codes written in Fortran, C, C++, Python, and MATLAB (sequential). PETSc provides many of the mechanisms neededmore » within parallel application codes, such as parallel matrix and vector assembly routines. The library is organized hierarchically, enabling users to employ the level of abstraction that is most appropriate for a particular problem. By using techniques of object-oriented programming, PETSc provides enormous flexibility for users. PETSc is a sophisticated set of software tools; as such, for some users it initially has a much steeper learning curve than a simple subroutine library. In particular, for individuals without some computer science background, experience programming in C, C++ or Fortran and experience using a debugger such as gdb or dbx, it may require a significant amount of time to take full advantage of the features that enable efficient software use. However, the power of the PETSc design and the algorithms it incorporates may make the efficient implementation of many application codes simpler than “rolling them” yourself. ;For many tasks a package such as MATLAB is often the best tool; PETSc is not intended for the classes of problems for which effective MATLAB code can be written. PETSc also has a MATLAB interface, so portions of your code can be written in MATLAB to “try out” the PETSc solvers. The resulting code will not be scalable however because currently MATLAB is inherently not scalable; and PETSc should not be used to attempt to provide a “parallel linear solver” in an otherwise sequential code. Certainly all parts of a previously sequential code need not be parallelized but the matrix generation portion must be parallelized to expect any kind of reasonable performance. Do not expect to generate your matrix sequentially and then “use PETSc” to solve the linear system in parallel. Since PETSc is under continued development, small changes in usage and calling sequences of routines will occur. PETSc is supported; see the web site http://www.mcs.anl.gov/petsc for information on contacting support. A http://www.mcs.anl.gov/petsc/publications may be found a list of publications and web sites that feature work involving PETSc. We welcome any reports of corrections for this document.« less

Adaptive particle-based pore-level modeling of incompressible fluid flow in porous media: a direct and parallel approach

NASA Astrophysics Data System (ADS)

Ovaysi, S.; Piri, M.

2009-12-01

We present a three-dimensional fully dynamic parallel particle-based model for direct pore-level simulation of incompressible viscous fluid flow in disordered porous media. The model was developed from scratch and is capable of simulating flow directly in three-dimensional high-resolution microtomography images of naturally occurring or man-made porous systems. It reads the images as input where the position of the solid walls are given. The entire medium, i.e., solid and fluid, is then discretized using particles. The model is based on Moving Particle Semi-implicit (MPS) technique. We modify this technique in order to improve its stability. The model handles highly irregular fluid-solid boundaries effectively. It takes into account viscous pressure drop in addition to the gravity forces. It conserves mass and can automatically detect any false connectivity with fluid particles in the neighboring pores and throats. It includes a sophisticated algorithm to automatically split and merge particles to maintain hydraulic connectivity of extremely narrow conduits. Furthermore, it uses novel methods to handle particle inconsistencies and open boundaries. To handle the computational load, we present a fully parallel version of the model that runs on distributed memory computer clusters and exhibits excellent scalability. The model is used to simulate unsteady-state flow problems under different conditions starting from straight noncircular capillary tubes with different cross-sectional shapes, i.e., circular/elliptical, square/rectangular and triangular cross-sections. We compare the predicted dimensionless hydraulic conductances with the data available in the literature and observe an excellent agreement. We then test the scalability of our parallel model with two samples of an artificial sandstone, samples A and B, with different volumes and different distributions (non-uniform and uniform) of solid particles among the processors. An excellent linear scalability is obtained for sample B that has more uniform distribution of solid particles leading to a superior load balancing. The model is then used to simulate fluid flow directly in REV size three-dimensional x-ray images of a naturally occurring sandstone. We analyze the quality and consistency of the predicted flow behavior and calculate absolute permeability, which compares well with the available network modeling and Lattice-Boltzmann permeabilities available in the literature for the same sandstone. We show that the model conserves mass very well and is stable computationally even at very narrow fluid conduits. The transient- and the steady-state fluid flow patterns are presented as well as the steady-state flow rates to compute absolute permeability. Furthermore, we discuss the vital role of our adaptive particle resolution scheme in preserving the original pore connectivity of the samples and their narrow channels through splitting and merging of fluid particles.

AHPCRC (Army High Performance Computing Research Center) Bulletin. Volume 2, Issue 1

DTIC Science & Technology

2010-01-01

Researchers in AHPCRC Technical Area 4 focus on improving processes for developing scalable, accurate parallel programs that are easily ported from one...control number. 1. REPORT DATE 2011 2. REPORT TYPE 3. DATES COVERED 00-00-2011 to 00-00-2011 4 . TITLE AND SUBTITLE AHPCRC (Army High...continued on page 4 Virtual levels in Sequoia represent an abstract memory hierarchy without specifying data transfer mechanisms, giving the

A task-based parallelism and vectorized approach to 3D Method of Characteristics (MOC) reactor simulation for high performance computing architectures

NASA Astrophysics Data System (ADS)

Tramm, John R.; Gunow, Geoffrey; He, Tim; Smith, Kord S.; Forget, Benoit; Siegel, Andrew R.

2016-05-01

In this study we present and analyze a formulation of the 3D Method of Characteristics (MOC) technique applied to the simulation of full core nuclear reactors. Key features of the algorithm include a task-based parallelism model that allows independent MOC tracks to be assigned to threads dynamically, ensuring load balancing, and a wide vectorizable inner loop that takes advantage of modern SIMD computer architectures. The algorithm is implemented in a set of highly optimized proxy applications in order to investigate its performance characteristics on CPU, GPU, and Intel Xeon Phi architectures. Speed, power, and hardware cost efficiencies are compared. Additionally, performance bottlenecks are identified for each architecture in order to determine the prospects for continued scalability of the algorithm on next generation HPC architectures.

A high performance parallel computing architecture for robust image features

NASA Astrophysics Data System (ADS)

Zhou, Renyan; Liu, Leibo; Wei, Shaojun

2014-03-01

A design of parallel architecture for image feature detection and description is proposed in this article. The major component of this architecture is a 2D cellular network composed of simple reprogrammable processors, enabling the Hessian Blob Detector and Haar Response Calculation, which are the most computing-intensive stage of the Speeded Up Robust Features (SURF) algorithm. Combining this 2D cellular network and dedicated hardware for SURF descriptors, this architecture achieves real-time image feature detection with minimal software in the host processor. A prototype FPGA implementation of the proposed architecture achieves 1318.9 GOPS general pixel processing @ 100 MHz clock and achieves up to 118 fps in VGA (640 × 480) image feature detection. The proposed architecture is stand-alone and scalable so it is easy to be migrated into VLSI implementation.

BCYCLIC: A parallel block tridiagonal matrix cyclic solver

NASA Astrophysics Data System (ADS)

Hirshman, S. P.; Perumalla, K. S.; Lynch, V. E.; Sanchez, R.

2010-09-01

A block tridiagonal matrix is factored with minimal fill-in using a cyclic reduction algorithm that is easily parallelized. Storage of the factored blocks allows the application of the inverse to multiple right-hand sides which may not be known at factorization time. Scalability with the number of block rows is achieved with cyclic reduction, while scalability with the block size is achieved using multithreaded routines (OpenMP, GotoBLAS) for block matrix manipulation. This dual scalability is a noteworthy feature of this new solver, as well as its ability to efficiently handle arbitrary (non-powers-of-2) block row and processor numbers. Comparison with a state-of-the art parallel sparse solver is presented. It is expected that this new solver will allow many physical applications to optimally use the parallel resources on current supercomputers. Example usage of the solver in magneto-hydrodynamic (MHD), three-dimensional equilibrium solvers for high-temperature fusion plasmas is cited.

Mesh-free data transfer algorithms for partitioned multiphysics problems: Conservation, accuracy, and parallelism

DOE PAGES

Slattery, Stuart R.

2015-12-02

In this study we analyze and extend mesh-free algorithms for three-dimensional data transfer problems in partitioned multiphysics simulations. We first provide a direct comparison between a mesh-based weighted residual method using the common-refinement scheme and two mesh-free algorithms leveraging compactly supported radial basis functions: one using a spline interpolation and one using a moving least square reconstruction. Through the comparison we assess both the conservation and accuracy of the data transfer obtained from each of the methods. We do so for a varying set of geometries with and without curvature and sharp features and for functions with and without smoothnessmore » and with varying gradients. Our results show that the mesh-based and mesh-free algorithms are complementary with cases where each was demonstrated to perform better than the other. We then focus on the mesh-free methods by developing a set of algorithms to parallelize them based on sparse linear algebra techniques. This includes a discussion of fast parallel radius searching in point clouds and restructuring the interpolation algorithms to leverage data structures and linear algebra services designed for large distributed computing environments. The scalability of our new algorithms is demonstrated on a leadership class computing facility using a set of basic scaling studies. Finally, these scaling studies show that for problems with reasonable load balance, our new algorithms for both spline interpolation and moving least square reconstruction demonstrate both strong and weak scalability using more than 100,000 MPI processes with billions of degrees of freedom in the data transfer operation.« less

Parallel grid library for rapid and flexible simulation development

NASA Astrophysics Data System (ADS)

Honkonen, I.; von Alfthan, S.; Sandroos, A.; Janhunen, P.; Palmroth, M.

2013-04-01

We present an easy to use and flexible grid library for developing highly scalable parallel simulations. The distributed cartesian cell-refinable grid (dccrg) supports adaptive mesh refinement and allows an arbitrary C++ class to be used as cell data. The amount of data in grid cells can vary both in space and time allowing dccrg to be used in very different types of simulations, for example in fluid and particle codes. Dccrg transfers the data between neighboring cells on different processes transparently and asynchronously allowing one to overlap computation and communication. This enables excellent scalability at least up to 32 k cores in magnetohydrodynamic tests depending on the problem and hardware. In the version of dccrg presented here part of the mesh metadata is replicated between MPI processes reducing the scalability of adaptive mesh refinement (AMR) to between 200 and 600 processes. Dccrg is free software that anyone can use, study and modify and is available at https://gitorious.org/dccrg. Users are also kindly requested to cite this work when publishing results obtained with dccrg. Catalogue identifier: AEOM_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEOM_v1_0.html Program obtainable from: CPC Program Library, Queen’s University, Belfast, N. Ireland Licensing provisions: GNU Lesser General Public License version 3 No. of lines in distributed program, including test data, etc.: 54975 No. of bytes in distributed program, including test data, etc.: 974015 Distribution format: tar.gz Programming language: C++. Computer: PC, cluster, supercomputer. Operating system: POSIX. The code has been parallelized using MPI and tested with 1-32768 processes RAM: 10 MB-10 GB per process Classification: 4.12, 4.14, 6.5, 19.3, 19.10, 20. External routines: MPI-2 [1], boost [2], Zoltan [3], sfc++ [4] Nature of problem: Grid library supporting arbitrary data in grid cells, parallel adaptive mesh refinement, transparent remote neighbor data updates and load balancing. Solution method: The simulation grid is represented by an adjacency list (graph) with vertices stored into a hash table and edges into contiguous arrays. Message Passing Interface standard is used for parallelization. Cell data is given as a template parameter when instantiating the grid. Restrictions: Logically cartesian grid. Running time: Running time depends on the hardware, problem and the solution method. Small problems can be solved in under a minute and very large problems can take weeks. The examples and tests provided with the package take less than about one minute using default options. In the version of dccrg presented here the speed of adaptive mesh refinement is at most of the order of 106 total created cells per second. http://www.mpi-forum.org/. http://www.boost.org/. K. Devine, E. Boman, R. Heaphy, B. Hendrickson, C. Vaughan, Zoltan data management services for parallel dynamic applications, Comput. Sci. Eng. 4 (2002) 90-97. http://dx.doi.org/10.1109/5992.988653. https://gitorious.org/sfc++.

A Parallel Numerical Micromagnetic Code Using FEniCS

NASA Astrophysics Data System (ADS)

Nagy, L.; Williams, W.; Mitchell, L.

2013-12-01

Many problems in the geosciences depend on understanding the ability of magnetic minerals to provide stable paleomagnetic recordings. Numerical micromagnetic modelling allows us to calculate the domain structures found in naturally occurring magnetic materials. However the computational cost rises exceedingly quickly with respect to the size and complexity of the geometries that we wish to model. This problem is compounded by the fact that the modern processor design no longer focuses on the speed at which calculations are performed, but rather on the number of computational units amongst which we may distribute our calculations. Consequently to better exploit modern computational resources our micromagnetic simulations must "go parallel". We present a parallel and scalable micromagnetics code written using FEniCS. FEniCS is a multinational collaboration involving several institutions (University of Cambridge, University of Chicago, The Simula Research Laboratory, etc.) that aims to provide a set of tools for writing scientific software; in particular software that employs the finite element method. The advantages of this approach are the leveraging of pre-existing projects from the world of scientific computing (PETSc, Trilinos, Metis/Parmetis, etc.) and exposing these so that researchers may pose problems in a manner closer to the mathematical language of their domain. Our code provides a scriptable interface (in Python) that allows users to not only run micromagnetic models in parallel, but also to perform pre/post processing of data.

The portals 4.0.1 network programming interface.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Barrett, Brian W.; Brightwell, Ronald Brian; Pedretti, Kevin

2013-04-01

This report presents a specification for the Portals 4.0 network programming interface. Portals 4.0 is intended to allow scalable, high-performance network communication between nodes of a parallel computing system. Portals 4.0 is well suited to massively parallel processing and embedded systems. Portals 4.0 represents an adaption of the data movement layer developed for massively parallel processing platforms, such as the 4500-node Intel TeraFLOPS machine. Sandias Cplant cluster project motivated the development of Version 3.0, which was later extended to Version 3.3 as part of the Cray Red Storm machine and XT line. Version 4.0 is targeted to the next generationmore » of machines employing advanced network interface architectures that support enhanced offload capabilities. 3« less

Revisiting Parallel Cyclic Reduction and Parallel Prefix-Based Algorithms for Block Tridiagonal System of Equations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Seal, Sudip K; Perumalla, Kalyan S; Hirshman, Steven Paul

2013-01-01

Simulations that require solutions of block tridiagonal systems of equations rely on fast parallel solvers for runtime efficiency. Leading parallel solvers that are highly effective for general systems of equations, dense or sparse, are limited in scalability when applied to block tridiagonal systems. This paper presents scalability results as well as detailed analyses of two parallel solvers that exploit the special structure of block tridiagonal matrices to deliver superior performance, often by orders of magnitude. A rigorous analysis of their relative parallel runtimes is shown to reveal the existence of a critical block size that separates the parameter space spannedmore » by the number of block rows, the block size and the processor count, into distinct regions that favor one or the other of the two solvers. Dependence of this critical block size on the above parameters as well as on machine-specific constants is established. These formal insights are supported by empirical results on up to 2,048 cores of a Cray XT4 system. To the best of our knowledge, this is the highest reported scalability for parallel block tridiagonal solvers to date.« less

A novel approach to multiple sequence alignment using hadoop data grids.

PubMed

Sudha Sadasivam, G; Baktavatchalam, G

2010-01-01

Multiple alignment of protein sequences helps to determine evolutionary linkage and to predict molecular structures. The factors to be considered while aligning multiple sequences are speed and accuracy of alignment. Although dynamic programming algorithms produce accurate alignments, they are computation intensive. In this paper we propose a time efficient approach to sequence alignment that also produces quality alignment. The dynamic nature of the algorithm coupled with data and computational parallelism of hadoop data grids improves the accuracy and speed of sequence alignment. The principle of block splitting in hadoop coupled with its scalability facilitates alignment of very large sequences.

Ultrahigh-order Maxwell solver with extreme scalability for electromagnetic PIC simulations of plasmas

NASA Astrophysics Data System (ADS)

Vincenti, Henri; Vay, Jean-Luc

2018-07-01

The advent of massively parallel supercomputers, with their distributed-memory technology using many processing units, has favored the development of highly-scalable local low-order solvers at the expense of harder-to-scale global very high-order spectral methods. Indeed, FFT-based methods, which were very popular on shared memory computers, have been largely replaced by finite-difference (FD) methods for the solution of many problems, including plasmas simulations with electromagnetic Particle-In-Cell methods. For some problems, such as the modeling of so-called "plasma mirrors" for the generation of high-energy particles and ultra-short radiations, we have shown that the inaccuracies of standard FD-based PIC methods prevent the modeling on present supercomputers at sufficient accuracy. We demonstrate here that a new method, based on the use of local FFTs, enables ultrahigh-order accuracy with unprecedented scalability, and thus for the first time the accurate modeling of plasma mirrors in 3D.

Effects of Ordering Strategies and Programming Paradigms on Sparse Matrix Computations

NASA Technical Reports Server (NTRS)

Oliker, Leonid; Li, Xiaoye; Husbands, Parry; Biswas, Rupak; Biegel, Bryan (Technical Monitor)

2002-01-01

The Conjugate Gradient (CG) algorithm is perhaps the best-known iterative technique to solve sparse linear systems that are symmetric and positive definite. For systems that are ill-conditioned, it is often necessary to use a preconditioning technique. In this paper, we investigate the effects of various ordering and partitioning strategies on the performance of parallel CG and ILU(O) preconditioned CG (PCG) using different programming paradigms and architectures. Results show that for this class of applications: ordering significantly improves overall performance on both distributed and distributed shared-memory systems, that cache reuse may be more important than reducing communication, that it is possible to achieve message-passing performance using shared-memory constructs through careful data ordering and distribution, and that a hybrid MPI+OpenMP paradigm increases programming complexity with little performance gains. A implementation of CG on the Cray MTA does not require special ordering or partitioning to obtain high efficiency and scalability, giving it a distinct advantage for adaptive applications; however, it shows limited scalability for PCG due to a lack of thread level parallelism.

Scalable Nonlinear Solvers for Fully Implicit Coupled Nuclear Fuel Modeling. Final Report

DOE Office of Scientific and Technical Information (OSTI.GOV)

Cai, Xiao-Chuan; Keyes, David; Yang, Chao

2014-09-29

The focus of the project is on the development and customization of some highly scalable domain decomposition based preconditioning techniques for the numerical solution of nonlinear, coupled systems of partial differential equations (PDEs) arising from nuclear fuel simulations. These high-order PDEs represent multiple interacting physical fields (for example, heat conduction, oxygen transport, solid deformation), each is modeled by a certain type of Cahn-Hilliard and/or Allen-Cahn equations. Most existing approaches involve a careful splitting of the fields and the use of field-by-field iterations to obtain a solution of the coupled problem. Such approaches have many advantages such as ease of implementationmore » since only single field solvers are needed, but also exhibit disadvantages. For example, certain nonlinear interactions between the fields may not be fully captured, and for unsteady problems, stable time integration schemes are difficult to design. In addition, when implemented on large scale parallel computers, the sequential nature of the field-by-field iterations substantially reduces the parallel efficiency. To overcome the disadvantages, fully coupled approaches have been investigated in order to obtain full physics simulations.« less

3D Data Denoising via Nonlocal Means Filter by Using Parallel GPU Strategies

PubMed Central

Cuomo, Salvatore; De Michele, Pasquale; Piccialli, Francesco

2014-01-01

Nonlocal Means (NLM) algorithm is widely considered as a state-of-the-art denoising filter in many research fields. Its high computational complexity leads researchers to the development of parallel programming approaches and the use of massively parallel architectures such as the GPUs. In the recent years, the GPU devices had led to achieving reasonable running times by filtering, slice-by-slice, and 3D datasets with a 2D NLM algorithm. In our approach we design and implement a fully 3D NonLocal Means parallel approach, adopting different algorithm mapping strategies on GPU architecture and multi-GPU framework, in order to demonstrate its high applicability and scalability. The experimental results we obtained encourage the usability of our approach in a large spectrum of applicative scenarios such as magnetic resonance imaging (MRI) or video sequence denoising. PMID:25045397

Parallel processor for real-time structural control

DOE Office of Scientific and Technical Information (OSTI.GOV)

Tise, B.L.

1992-01-01

A parallel processor that is optimized for real-time linear control has been developed. This modular system consists of A/D modules, D/A modules, and floating-point processor modules. The scalable processor uses up to 1,000 Motorola DSP96002 floating-point processors for a peak computational rate of 60 GFLOPS. Sampling rates up to 625 kHz are supported by this analog-in to analog-out controller. The high processing rate and parallel architecture make this processor suitable for computing state-space equations and other multiply/accumulate-intensive digital filters. Processor features include 14-bit conversion devices, low input-output latency, 240 Mbyte/s synchronous backplane bus, low-skew clock distribution circuit, VME connection tomore » host computer, parallelizing code generator, and look-up-tables for actuator linearization. This processor was designed primarily for experiments in structural control. The A/D modules sample sensors mounted on the structure and the floating-point processor modules compute the outputs using the programmed control equations. The outputs are sent through the D/A module to the power amps used to drive the structure's actuators. The host computer is a Sun workstation. An Open Windows-based control panel is provided to facilitate data transfer to and from the processor, as well as to control the operating mode of the processor. A diagnostic mode is provided to allow stimulation of the structure and acquisition of the structural response via sensor inputs.« less

McrEngine: A Scalable Checkpointing System Using Data-Aware Aggregation and Compression

DOE PAGES

Islam, Tanzima Zerin; Mohror, Kathryn; Bagchi, Saurabh; ...

2013-01-01

High performance computing (HPC) systems use checkpoint-restart to tolerate failures. Typically, applications store their states in checkpoints on a parallel file system (PFS). As applications scale up, checkpoint-restart incurs high overheads due to contention for PFS resources. The high overheads force large-scale applications to reduce checkpoint frequency, which means more compute time is lost in the event of failure. We alleviate this problem through a scalable checkpoint-restart system, mcrEngine. McrEngine aggregates checkpoints from multiple application processes with knowledge of the data semantics available through widely-used I/O libraries, e.g., HDF5 and netCDF, and compresses them. Our novel scheme improves compressibility ofmore » checkpoints up to 115% over simple concatenation and compression. Our evaluation with large-scale application checkpoints show that mcrEngine reduces checkpointing overhead by up to 87% and restart overhead by up to 62% over a baseline with no aggregation or compression.« less

Extreme-Scale Stochastic Particle Tracing for Uncertain Unsteady Flow Analysis

DOE Office of Scientific and Technical Information (OSTI.GOV)

Guo, Hanqi; He, Wenbin; Seo, Sangmin

2016-11-13

We present an efficient and scalable solution to estimate uncertain transport behaviors using stochastic flow maps (SFM,) for visualizing and analyzing uncertain unsteady flows. SFM computation is extremely expensive because it requires many Monte Carlo runs to trace densely seeded particles in the flow. We alleviate the computational cost by decoupling the time dependencies in SFMs so that we can process adjacent time steps independently and then compose them together for longer time periods. Adaptive refinement is also used to reduce the number of runs for each location. We then parallelize over tasks—packets of particles in our design—to achieve highmore » efficiency in MPI/thread hybrid programming. Such a task model also enables CPU/GPU coprocessing. We show the scalability on two supercomputers, Mira (up to 1M Blue Gene/Q cores) and Titan (up to 128K Opteron cores and 8K GPUs), that can trace billions of particles in seconds.« less

Scalable designs for quasiparticle-poisoning-protected topological quantum computation with Majorana zero modes

NASA Astrophysics Data System (ADS)

Karzig, Torsten; Knapp, Christina; Lutchyn, Roman M.; Bonderson, Parsa; Hastings, Matthew B.; Nayak, Chetan; Alicea, Jason; Flensberg, Karsten; Plugge, Stephan; Oreg, Yuval; Marcus, Charles M.; Freedman, Michael H.

2017-06-01

We present designs for scalable quantum computers composed of qubits encoded in aggregates of four or more Majorana zero modes, realized at the ends of topological superconducting wire segments that are assembled into superconducting islands with significant charging energy. Quantum information can be manipulated according to a measurement-only protocol, which is facilitated by tunable couplings between Majorana zero modes and nearby semiconductor quantum dots. Our proposed architecture designs have the following principal virtues: (1) the magnetic field can be aligned in the direction of all of the topological superconducting wires since they are all parallel; (2) topological T junctions are not used, obviating possible difficulties in their fabrication and utilization; (3) quasiparticle poisoning is abated by the charging energy; (4) Clifford operations are executed by a relatively standard measurement: detection of corrections to quantum dot energy, charge, or differential capacitance induced by quantum fluctuations; (5) it is compatible with strategies for producing good approximate magic states.

Parallel implementation of the particle simulation method with dynamic load balancing: Toward realistic geodynamical simulation

NASA Astrophysics Data System (ADS)

Furuichi, M.; Nishiura, D.

2015-12-01

Fully Lagrangian methods such as Smoothed Particle Hydrodynamics (SPH) and Discrete Element Method (DEM) have been widely used to solve the continuum and particles motions in the computational geodynamics field. These mesh-free methods are suitable for the problems with the complex geometry and boundary. In addition, their Lagrangian nature allows non-diffusive advection useful for tracking history dependent properties (e.g. rheology) of the material. These potential advantages over the mesh-based methods offer effective numerical applications to the geophysical flow and tectonic processes, which are for example, tsunami with free surface and floating body, magma intrusion with fracture of rock, and shear zone pattern generation of granular deformation. In order to investigate such geodynamical problems with the particle based methods, over millions to billion particles are required for the realistic simulation. Parallel computing is therefore important for handling such huge computational cost. An efficient parallel implementation of SPH and DEM methods is however known to be difficult especially for the distributed-memory architecture. Lagrangian methods inherently show workload imbalance problem for parallelization with the fixed domain in space, because particles move around and workloads change during the simulation. Therefore dynamic load balance is key technique to perform the large scale SPH and DEM simulation. In this work, we present the parallel implementation technique of SPH and DEM method utilizing dynamic load balancing algorithms toward the high resolution simulation over large domain using the massively parallel super computer system. Our method utilizes the imbalances of the executed time of each MPI process as the nonlinear term of parallel domain decomposition and minimizes them with the Newton like iteration method. In order to perform flexible domain decomposition in space, the slice-grid algorithm is used. Numerical tests show that our approach is suitable for solving the particles with different calculation costs (e.g. boundary particles) as well as the heterogeneous computer architecture. We analyze the parallel efficiency and scalability on the super computer systems (K-computer, Earth simulator 3, etc.).

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lusk, Ewing; Butler, Ralph; Pieper, Steven C.

Here, we take a historical approach to our presentation of self-scheduled task parallelism, a programming model with its origins in early irregular and nondeterministic computations encountered in automated theorem proving and logic programming. We show how an extremely simple task model has evolved into a system, asynchronous dynamic load balancing (ADLB), and a scalable implementation capable of supporting sophisticated applications on today’s (and tomorrow’s) largest supercomputers; and we illustrate the use of ADLB with a Green’s function Monte Carlo application, a modern, mature nuclear physics code in production use. Our lesson is that by surrendering a certain amount of generalitymore » and thus applicability, a minimal programming model (in terms of its basic concepts and the size of its application programmer interface) can achieve extreme scalability without introducing complexity.« less

A complexity-scalable software-based MPEG-2 video encoder.

PubMed

Chen, Guo-bin; Lu, Xin-ning; Wang, Xing-guo; Liu, Ji-lin

2004-05-01

With the development of general-purpose processors (GPP) and video signal processing algorithms, it is possible to implement a software-based real-time video encoder on GPP, and its low cost and easy upgrade attract developers' interests to transfer video encoding from specialized hardware to more flexible software. In this paper, the encoding structure is set up first to support complexity scalability; then a lot of high performance algorithms are used on the key time-consuming modules in coding process; finally, at programming level, processor characteristics are considered to improve data access efficiency and processing parallelism. Other programming methods such as lookup table are adopted to reduce the computational complexity. Simulation results showed that these ideas could not only improve the global performance of video coding, but also provide great flexibility in complexity regulation.

Extreme Performance Scalable Operating Systems Final Progress Report (July 1, 2008 - October 31, 2011)

DOE Office of Scientific and Technical Information (OSTI.GOV)

Malony, Allen D; Shende, Sameer

This is the final progress report for the FastOS (Phase 2) (FastOS-2) project with Argonne National Laboratory and the University of Oregon (UO). The project started at UO on July 1, 2008 and ran until April 30, 2010, at which time a six-month no-cost extension began. The FastOS-2 work at UO delivered excellent results in all research work areas: * scalable parallel monitoring * kernel-level performance measurement * parallel I/0 system measurement * large-scale and hybrid application performance measurement * onlne scalable performance data reduction and analysis * binary instrumentation

Development and application of numerical techniques for general-relativistic magnetohydrodynamics simulations of black hole accretion

NASA Astrophysics Data System (ADS)

White, Christopher Joseph

We describe the implementation of sophisticated numerical techniques for general-relativistic magnetohydrodynamics simulations in the Athena++ code framework. Improvements over many existing codes include the use of advanced Riemann solvers and of staggered-mesh constrained transport. Combined with considerations for computational performance and parallel scalability, these allow us to investigate black hole accretion flows with unprecedented accuracy. The capability of the code is demonstrated by exploring magnetically arrested disks.

SAChES: Scalable Adaptive Chain-Ensemble Sampling.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Swiler, Laura Painton; Ray, Jaideep; Ebeida, Mohamed Salah

We present the development of a parallel Markov Chain Monte Carlo (MCMC) method called SAChES, Scalable Adaptive Chain-Ensemble Sampling. This capability is targed to Bayesian calibration of com- putationally expensive simulation models. SAChES involves a hybrid of two methods: Differential Evo- lution Monte Carlo followed by Adaptive Metropolis. Both methods involve parallel chains. Differential evolution allows one to explore high-dimensional parameter spaces using loosely coupled (i.e., largely asynchronous) chains. Loose coupling allows the use of large chain ensembles, with far more chains than the number of parameters to explore. This reduces per-chain sampling burden, enables high-dimensional inversions and the usemore » of computationally expensive forward models. The large number of chains can also ameliorate the impact of silent-errors, which may affect only a few chains. The chain ensemble can also be sampled to provide an initial condition when an aberrant chain is re-spawned. Adaptive Metropolis takes the best points from the differential evolution and efficiently hones in on the poste- rior density. The multitude of chains in SAChES is leveraged to (1) enable efficient exploration of the parameter space; and (2) ensure robustness to silent errors which may be unavoidable in extreme-scale computational platforms of the future. This report outlines SAChES, describes four papers that are the result of the project, and discusses some additional results.« less

Global interrupt and barrier networks

DOEpatents

Blumrich, Matthias A.; Chen, Dong; Coteus, Paul W.; Gara, Alan G.; Giampapa, Mark E; Heidelberger, Philip; Kopcsay, Gerard V.; Steinmacher-Burow, Burkhard D.; Takken, Todd E.

2008-10-28

A system and method for generating global asynchronous signals in a computing structure. Particularly, a global interrupt and barrier network is implemented that implements logic for generating global interrupt and barrier signals for controlling global asynchronous operations performed by processing elements at selected processing nodes of a computing structure in accordance with a processing algorithm; and includes the physical interconnecting of the processing nodes for communicating the global interrupt and barrier signals to the elements via low-latency paths. The global asynchronous signals respectively initiate interrupt and barrier operations at the processing nodes at times selected for optimizing performance of the processing algorithms. In one embodiment, the global interrupt and barrier network is implemented in a scalable, massively parallel supercomputing device structure comprising a plurality of processing nodes interconnected by multiple independent networks, with each node including one or more processing elements for performing computation or communication activity as required when performing parallel algorithm operations. One multiple independent network includes a global tree network for enabling high-speed global tree communications among global tree network nodes or sub-trees thereof. The global interrupt and barrier network may operate in parallel with the global tree network for providing global asynchronous sideband signals.

Serial Back-Plane Technologies in Advanced Avionics Architectures

NASA Technical Reports Server (NTRS)

Varnavas, Kosta

2005-01-01

Current back plane technologies such as VME, and current personal computer back planes such as PCI, are shared bus systems that can exhibit nondeterministic latencies. This means a card can take control of the bus and use resources indefinitely affecting the ability of other cards in the back plane to acquire the bus. This provides a real hit on the reliability of the system. Additionally, these parallel busses only have bandwidths in the 100s of megahertz range and EMI and noise effects get worse the higher the bandwidth goes. To provide scalable, fault-tolerant, advanced computing systems, more applicable to today s connected computing environment and to better meet the needs of future requirements for advanced space instruments and vehicles, serial back-plane technologies should be implemented in advanced avionics architectures. Serial backplane technologies eliminate the problem of one card getting the bus and never relinquishing it, or one minor problem on the backplane bringing the whole system down. Being serial instead of parallel improves the reliability by reducing many of the signal integrity issues associated with parallel back planes and thus significantly improves reliability. The increased speeds associated with a serial backplane are an added bonus.

Parallel implementation of a Lagrangian-based model on an adaptive mesh in C++: Application to sea-ice

NASA Astrophysics Data System (ADS)

Samaké, Abdoulaye; Rampal, Pierre; Bouillon, Sylvain; Ólason, Einar

2017-12-01

We present a parallel implementation framework for a new dynamic/thermodynamic sea-ice model, called neXtSIM, based on the Elasto-Brittle rheology and using an adaptive mesh. The spatial discretisation of the model is done using the finite-element method. The temporal discretisation is semi-implicit and the advection is achieved using either a pure Lagrangian scheme or an Arbitrary Lagrangian Eulerian scheme (ALE). The parallel implementation presented here focuses on the distributed-memory approach using the message-passing library MPI. The efficiency and the scalability of the parallel algorithms are illustrated by the numerical experiments performed using up to 500 processor cores of a cluster computing system. The performance obtained by the proposed parallel implementation of the neXtSIM code is shown being sufficient to perform simulations for state-of-the-art sea ice forecasting and geophysical process studies over geographical domain of several millions squared kilometers like the Arctic region.

Homemade Buckeye-Pi: A Learning Many-Node Platform for High-Performance Parallel Computing

NASA Astrophysics Data System (ADS)

Amooie, M. A.; Moortgat, J.

2017-12-01

We report on the "Buckeye-Pi" cluster, the supercomputer developed in The Ohio State University School of Earth Sciences from 128 inexpensive Raspberry Pi (RPi) 3 Model B single-board computers. Each RPi is equipped with fast Quad Core 1.2GHz ARMv8 64bit processor, 1GB of RAM, and 32GB microSD card for local storage. Therefore, the cluster has a total RAM of 128GB that is distributed on the individual nodes and a flash capacity of 4TB with 512 processors, while it benefits from low power consumption, easy portability, and low total cost. The cluster uses the Message Passing Interface protocol to manage the communications between each node. These features render our platform the most powerful RPi supercomputer to date and suitable for educational applications in high-performance-computing (HPC) and handling of large datasets. In particular, we use the Buckeye-Pi to implement optimized parallel codes in our in-house simulator for subsurface media flows with the goal of achieving a massively-parallelized scalable code. We present benchmarking results for the computational performance across various number of RPi nodes. We believe our project could inspire scientists and students to consider the proposed unconventional cluster architecture as a mainstream and a feasible learning platform for challenging engineering and scientific problems.

Shared virtual memory and generalized speedup

NASA Technical Reports Server (NTRS)

Sun, Xian-He; Zhu, Jianping

1994-01-01

Generalized speedup is defined as parallel speed over sequential speed. The generalized speedup and its relation with other existing performance metrics, such as traditional speedup, efficiency, scalability, etc., are carefully studied. In terms of the introduced asymptotic speed, it was shown that the difference between the generalized speedup and the traditional speedup lies in the definition of the efficiency of uniprocessor processing, which is a very important issue in shared virtual memory machines. A scientific application was implemented on a KSR-1 parallel computer. Experimental and theoretical results show that the generalized speedup is distinct from the traditional speedup and provides a more reasonable measurement. In the study of different speedups, various causes of superlinear speedup are also presented.

Parallel Transport Quantum Logic Gates with Trapped Ions.

PubMed

de Clercq, Ludwig E; Lo, Hsiang-Yu; Marinelli, Matteo; Nadlinger, David; Oswald, Robin; Negnevitsky, Vlad; Kienzler, Daniel; Keitch, Ben; Home, Jonathan P

2016-02-26

We demonstrate single-qubit operations by transporting a beryllium ion with a controlled velocity through a stationary laser beam. We use these to perform coherent sequences of quantum operations, and to perform parallel quantum logic gates on two ions in different processing zones of a multiplexed ion trap chip using a single recycled laser beam. For the latter, we demonstrate individually addressed single-qubit gates by local control of the speed of each ion. The fidelities we observe are consistent with operations performed using standard methods involving static ions and pulsed laser fields. This work therefore provides a path to scalable ion trap quantum computing with reduced requirements on the optical control complexity.

Scalable computing for evolutionary genomics.

PubMed

Prins, Pjotr; Belhachemi, Dominique; Möller, Steffen; Smant, Geert

2012-01-01

Genomic data analysis in evolutionary biology is becoming so computationally intensive that analysis of multiple hypotheses and scenarios takes too long on a single desktop computer. In this chapter, we discuss techniques for scaling computations through parallelization of calculations, after giving a quick overview of advanced programming techniques. Unfortunately, parallel programming is difficult and requires special software design. The alternative, especially attractive for legacy software, is to introduce poor man's parallelization by running whole programs in parallel as separate processes, using job schedulers. Such pipelines are often deployed on bioinformatics computer clusters. Recent advances in PC virtualization have made it possible to run a full computer operating system, with all of its installed software, on top of another operating system, inside a "box," or virtual machine (VM). Such a VM can flexibly be deployed on multiple computers, in a local network, e.g., on existing desktop PCs, and even in the Cloud, to create a "virtual" computer cluster. Many bioinformatics applications in evolutionary biology can be run in parallel, running processes in one or more VMs. Here, we show how a ready-made bioinformatics VM image, named BioNode, effectively creates a computing cluster, and pipeline, in a few steps. This allows researchers to scale-up computations from their desktop, using available hardware, anytime it is required. BioNode is based on Debian Linux and can run on networked PCs and in the Cloud. Over 200 bioinformatics and statistical software packages, of interest to evolutionary biology, are included, such as PAML, Muscle, MAFFT, MrBayes, and BLAST. Most of these software packages are maintained through the Debian Med project. In addition, BioNode contains convenient configuration scripts for parallelizing bioinformatics software. Where Debian Med encourages packaging free and open source bioinformatics software through one central project, BioNode encourages creating free and open source VM images, for multiple targets, through one central project. BioNode can be deployed on Windows, OSX, Linux, and in the Cloud. Next to the downloadable BioNode images, we provide tutorials online, which empower bioinformaticians to install and run BioNode in different environments, as well as information for future initiatives, on creating and building such images.

Unstructured Adaptive Grid Computations on an Array of SMPs

NASA Technical Reports Server (NTRS)

Biswas, Rupak; Pramanick, Ira; Sohn, Andrew; Simon, Horst D.

1996-01-01

Dynamic load balancing is necessary for parallel adaptive methods to solve unsteady CFD problems on unstructured grids. We have presented such a dynamic load balancing framework called JOVE, in this paper. Results on a four-POWERnode POWER CHALLENGEarray demonstrated that load balancing gives significant performance improvements over no load balancing for such adaptive computations. The parallel speedup of JOVE, implemented using MPI on the POWER CHALLENCEarray, was significant, being as high as 31 for 32 processors. An implementation of JOVE that exploits 'an array of SMPS' architecture was also studied; this hybrid JOVE outperformed flat JOVE by up to 28% on the meshes and adaption models tested. With large, realistic meshes and actual flow-solver and adaption phases incorporated into JOVE, hybrid JOVE can be expected to yield significant advantage over flat JOVE, especially as the number of processors is increased, thus demonstrating the scalability of an array of SMPs architecture.

Rapid protein alignment in the cloud: HAMOND combines fast DIAMOND alignments with Hadoop parallelism.

PubMed

Yu, Jia; Blom, Jochen; Sczyrba, Alexander; Goesmann, Alexander

2017-09-10

The introduction of next generation sequencing has caused a steady increase in the amounts of data that have to be processed in modern life science. Sequence alignment plays a key role in the analysis of sequencing data e.g. within whole genome sequencing or metagenome projects. BLAST is a commonly used alignment tool that was the standard approach for more than two decades, but in the last years faster alternatives have been proposed including RapSearch, GHOSTX, and DIAMOND. Here we introduce HAMOND, an application that uses Apache Hadoop to parallelize DIAMOND computation in order to scale-out the calculation of alignments. HAMOND is fault tolerant and scalable by utilizing large cloud computing infrastructures like Amazon Web Services. HAMOND has been tested in comparative genomics analyses and showed promising results both in efficiency and accuracy. Copyright © 2017 The Author(s). Published by Elsevier B.V. All rights reserved.

A Parallel Multigrid Solver for Viscous Flows on Anisotropic Structured Grids

NASA Technical Reports Server (NTRS)

Prieto, Manuel; Montero, Ruben S.; Llorente, Ignacio M.; Bushnell, Dennis M. (Technical Monitor)

2001-01-01

This paper presents an efficient parallel multigrid solver for speeding up the computation of a 3-D model that treats the flow of a viscous fluid over a flat plate. The main interest of this simulation lies in exhibiting some basic difficulties that prevent optimal multigrid efficiencies from being achieved. As the computing platform, we have used Coral, a Beowulf-class system based on Intel Pentium processors and equipped with GigaNet cLAN and switched Fast Ethernet networks. Our study not only examines the scalability of the solver but also includes a performance evaluation of Coral where the investigated solver has been used to compare several of its design choices, namely, the interconnection network (GigaNet versus switched Fast-Ethernet) and the node configuration (dual nodes versus single nodes). As a reference, the performance results have been compared with those obtained with the NAS-MG benchmark.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Luszczek, Piotr R; Tomov, Stanimire Z; Dongarra, Jack J

We present an efficient and scalable programming model for the development of linear algebra in heterogeneous multi-coprocessor environments. The model incorporates some of the current best design and implementation practices for the heterogeneous acceleration of dense linear algebra (DLA). Examples are given as the basis for solving linear systems' algorithms - the LU, QR, and Cholesky factorizations. To generate the extreme level of parallelism needed for the efficient use of coprocessors, algorithms of interest are redesigned and then split into well-chosen computational tasks. The tasks execution is scheduled over the computational components of a hybrid system of multi-core CPUs andmore » coprocessors using a light-weight runtime system. The use of lightweight runtime systems keeps scheduling overhead low, while enabling the expression of parallelism through otherwise sequential code. This simplifies the development efforts and allows the exploration of the unique strengths of the various hardware components.« less

A scalable variational inequality approach for flow through porous media models with pressure-dependent viscosity

NASA Astrophysics Data System (ADS)

Mapakshi, N. K.; Chang, J.; Nakshatrala, K. B.

2018-04-01

Mathematical models for flow through porous media typically enjoy the so-called maximum principles, which place bounds on the pressure field. It is highly desirable to preserve these bounds on the pressure field in predictive numerical simulations, that is, one needs to satisfy discrete maximum principles (DMP). Unfortunately, many of the existing formulations for flow through porous media models do not satisfy DMP. This paper presents a robust, scalable numerical formulation based on variational inequalities (VI), to model non-linear flows through heterogeneous, anisotropic porous media without violating DMP. VI is an optimization technique that places bounds on the numerical solutions of partial differential equations. To crystallize the ideas, a modification to Darcy equations by taking into account pressure-dependent viscosity will be discretized using the lowest-order Raviart-Thomas (RT0) and Variational Multi-scale (VMS) finite element formulations. It will be shown that these formulations violate DMP, and, in fact, these violations increase with an increase in anisotropy. It will be shown that the proposed VI-based formulation provides a viable route to enforce DMP. Moreover, it will be shown that the proposed formulation is scalable, and can work with any numerical discretization and weak form. A series of numerical benchmark problems are solved to demonstrate the effects of heterogeneity, anisotropy and non-linearity on DMP violations under the two chosen formulations (RT0 and VMS), and that of non-linearity on solver convergence for the proposed VI-based formulation. Parallel scalability on modern computational platforms will be illustrated through strong-scaling studies, which will prove the efficiency of the proposed formulation in a parallel setting. Algorithmic scalability as the problem size is scaled up will be demonstrated through novel static-scaling studies. The performed static-scaling studies can serve as a guide for users to be able to select an appropriate discretization for a given problem size.

Accelerating Climate and Weather Simulations through Hybrid Computing

NASA Technical Reports Server (NTRS)

Zhou, Shujia; Cruz, Carlos; Duffy, Daniel; Tucker, Robert; Purcell, Mark

2011-01-01

Unconventional multi- and many-core processors (e.g. IBM (R) Cell B.E.(TM) and NVIDIA (R) GPU) have emerged as effective accelerators in trial climate and weather simulations. Yet these climate and weather models typically run on parallel computers with conventional processors (e.g. Intel, AMD, and IBM) using Message Passing Interface. To address challenges involved in efficiently and easily connecting accelerators to parallel computers, we investigated using IBM's Dynamic Application Virtualization (TM) (IBM DAV) software in a prototype hybrid computing system with representative climate and weather model components. The hybrid system comprises two Intel blades and two IBM QS22 Cell B.E. blades, connected with both InfiniBand(R) (IB) and 1-Gigabit Ethernet. The system significantly accelerates a solar radiation model component by offloading compute-intensive calculations to the Cell blades. Systematic tests show that IBM DAV can seamlessly offload compute-intensive calculations from Intel blades to Cell B.E. blades in a scalable, load-balanced manner. However, noticeable communication overhead was observed, mainly due to IP over the IB protocol. Full utilization of IB Sockets Direct Protocol and the lower latency production version of IBM DAV will reduce this overhead.

Optical interconnection networks for high-performance computing systems

NASA Astrophysics Data System (ADS)

Biberman, Aleksandr; Bergman, Keren

2012-04-01

Enabled by silicon photonic technology, optical interconnection networks have the potential to be a key disruptive technology in computing and communication industries. The enduring pursuit of performance gains in computing, combined with stringent power constraints, has fostered the ever-growing computational parallelism associated with chip multiprocessors, memory systems, high-performance computing systems and data centers. Sustaining these parallelism growths introduces unique challenges for on- and off-chip communications, shifting the focus toward novel and fundamentally different communication approaches. Chip-scale photonic interconnection networks, enabled by high-performance silicon photonic devices, offer unprecedented bandwidth scalability with reduced power consumption. We demonstrate that the silicon photonic platforms have already produced all the high-performance photonic devices required to realize these types of networks. Through extensive empirical characterization in much of our work, we demonstrate such feasibility of waveguides, modulators, switches and photodetectors. We also demonstrate systems that simultaneously combine many functionalities to achieve more complex building blocks. We propose novel silicon photonic devices, subsystems, network topologies and architectures to enable unprecedented performance of these photonic interconnection networks. Furthermore, the advantages of photonic interconnection networks extend far beyond the chip, offering advanced communication environments for memory systems, high-performance computing systems, and data centers.

Implementation of a Parallel Kalman Filter for Stratospheric Chemical Tracer Assimilation

NASA Technical Reports Server (NTRS)

Chang, Lang-Ping; Lyster, Peter M.; Menard, R.; Cohn, S. E.

1998-01-01

A Kalman filter for the assimilation of long-lived atmospheric chemical constituents has been developed for two-dimensional transport models on isentropic surfaces over the globe. An important attribute of the Kalman filter is that it calculates error covariances of the constituent fields using the tracer dynamics. Consequently, the current Kalman-filter assimilation is a five-dimensional problem (coordinates of two points and time), and it can only be handled on computers with large memory and high floating point speed. In this paper, an implementation of the Kalman filter for distributed-memory, message-passing parallel computers is discussed. Two approaches were studied: an operator decomposition and a covariance decomposition. The latter was found to be more scalable than the former, and it possesses the property that the dynamical model does not need to be parallelized, which is of considerable practical advantage. This code is currently used to assimilate constituent data retrieved by limb sounders on the Upper Atmosphere Research Satellite. Tests of the code examined the variance transport and observability properties. Aspects of the parallel implementation, some timing results, and a brief discussion of the physical results will be presented.

Recent Progress on the Parallel Implementation of Moving-Body Overset Grid Schemes

NASA Technical Reports Server (NTRS)

Wissink, Andrew; Allen, Edwin (Technical Monitor)

1998-01-01

Viscous calculations about geometrically complex bodies in which there is relative motion between component parts is one of the most computationally demanding problems facing CFD researchers today. This presentation documents results from the first two years of a CHSSI-funded effort within the U.S. Army AFDD to develop scalable dynamic overset grid methods for unsteady viscous calculations with moving-body problems. The first pan of the presentation will focus on results from OVERFLOW-D1, a parallelized moving-body overset grid scheme that employs traditional Chimera methodology. The two processes that dominate the cost of such problems are the flow solution on each component and the intergrid connectivity solution. Parallel implementations of the OVERFLOW flow solver and DCF3D connectivity software are coupled with a proposed two-part static-dynamic load balancing scheme and tested on the IBM SP and Cray T3E multi-processors. The second part of the presentation will cover some recent results from OVERFLOW-D2, a new flow solver that employs Cartesian grids with various levels of refinement, facilitating solution adaption. A study of the parallel performance of the scheme on large distributed- memory multiprocessor computer architectures will be reported.

Quantum simulator review

NASA Astrophysics Data System (ADS)

Bednar, Earl; Drager, Steven L.

2007-04-01

Quantum information processing's objective is to utilize revolutionary computing capability based on harnessing the paradigm shift offered by quantum computing to solve classically hard and computationally challenging problems. Some of our computationally challenging problems of interest include: the capability for rapid image processing, rapid optimization of logistics, protecting information, secure distributed simulation, and massively parallel computation. Currently, one important problem with quantum information processing is that the implementation of quantum computers is difficult to realize due to poor scalability and great presence of errors. Therefore, we have supported the development of Quantum eXpress and QuIDD Pro, two quantum computer simulators running on classical computers for the development and testing of new quantum algorithms and processes. This paper examines the different methods used by these two quantum computing simulators. It reviews both simulators, highlighting each simulators background, interface, and special features. It also demonstrates the implementation of current quantum algorithms on each simulator. It concludes with summary comments on both simulators.

Development, Verification and Validation of Parallel, Scalable Volume of Fluid CFD Program for Propulsion Applications

NASA Technical Reports Server (NTRS)

West, Jeff; Yang, H. Q.

2014-01-01

There are many instances involving liquid/gas interfaces and their dynamics in the design of liquid engine powered rockets such as the Space Launch System (SLS). Some examples of these applications are: Propellant tank draining and slosh, subcritical condition injector analysis for gas generators, preburners and thrust chambers, water deluge mitigation for launch induced environments and even solid rocket motor liquid slag dynamics. Commercially available CFD programs simulating gas/liquid interfaces using the Volume of Fluid approach are currently limited in their parallel scalability. In 2010 for instance, an internal NASA/MSFC review of three commercial tools revealed that parallel scalability was seriously compromised at 8 cpus and no additional speedup was possible after 32 cpus. Other non-interface CFD applications at the time were demonstrating useful parallel scalability up to 4,096 processors or more. Based on this review, NASA/MSFC initiated an effort to implement a Volume of Fluid implementation within the unstructured mesh, pressure-based algorithm CFD program, Loci-STREAM. After verification was achieved by comparing results to the commercial CFD program CFD-Ace+, and validation by direct comparison with data, Loci-STREAM-VoF is now the production CFD tool for propellant slosh force and slosh damping rate simulations at NASA/MSFC. On these applications, good parallel scalability has been demonstrated for problems sizes of tens of millions of cells and thousands of cpu cores. Ongoing efforts are focused on the application of Loci-STREAM-VoF to predict the transient flow patterns of water on the SLS Mobile Launch Platform in order to support the phasing of water for launch environment mitigation so that vehicle determinantal effects are not realized.

Development of a Robust and Efficient Parallel Solver for Unsteady Turbomachinery Flows

NASA Technical Reports Server (NTRS)

West, Jeff; Wright, Jeffrey; Thakur, Siddharth; Luke, Ed; Grinstead, Nathan

2012-01-01

The traditional design and analysis practice for advanced propulsion systems relies heavily on expensive full-scale prototype development and testing. Over the past decade, use of high-fidelity analysis and design tools such as CFD early in the product development cycle has been identified as one way to alleviate testing costs and to develop these devices better, faster and cheaper. In the design of advanced propulsion systems, CFD plays a major role in defining the required performance over the entire flight regime, as well as in testing the sensitivity of the design to the different modes of operation. Increased emphasis is being placed on developing and applying CFD models to simulate the flow field environments and performance of advanced propulsion systems. This necessitates the development of next generation computational tools which can be used effectively and reliably in a design environment. The turbomachinery simulation capability presented here is being developed in a computational tool called Loci-STREAM [1]. It integrates proven numerical methods for generalized grids and state-of-the-art physical models in a novel rule-based programming framework called Loci [2] which allows: (a) seamless integration of multidisciplinary physics in a unified manner, and (b) automatic handling of massively parallel computing. The objective is to be able to routinely simulate problems involving complex geometries requiring large unstructured grids and complex multidisciplinary physics. An immediate application of interest is simulation of unsteady flows in rocket turbopumps, particularly in cryogenic liquid rocket engines. The key components of the overall methodology presented in this paper are the following: (a) high fidelity unsteady simulation capability based on Detached Eddy Simulation (DES) in conjunction with second-order temporal discretization, (b) compliance with Geometric Conservation Law (GCL) in order to maintain conservative property on moving meshes for second-order time-stepping scheme, (c) a novel cloud-of-points interpolation method (based on a fast parallel kd-tree search algorithm) for interfaces between turbomachinery components in relative motion which is demonstrated to be highly scalable, and (d) demonstrated accuracy and parallel scalability on large grids (approx 250 million cells) in full turbomachinery geometries.

Hadoop-GIS: A High Performance Spatial Data Warehousing System over MapReduce.

PubMed

Aji, Ablimit; Wang, Fusheng; Vo, Hoang; Lee, Rubao; Liu, Qiaoling; Zhang, Xiaodong; Saltz, Joel

2013-08-01

Support of high performance queries on large volumes of spatial data becomes increasingly important in many application domains, including geospatial problems in numerous fields, location based services, and emerging scientific applications that are increasingly data- and compute-intensive. The emergence of massive scale spatial data is due to the proliferation of cost effective and ubiquitous positioning technologies, development of high resolution imaging technologies, and contribution from a large number of community users. There are two major challenges for managing and querying massive spatial data to support spatial queries: the explosion of spatial data, and the high computational complexity of spatial queries. In this paper, we present Hadoop-GIS - a scalable and high performance spatial data warehousing system for running large scale spatial queries on Hadoop. Hadoop-GIS supports multiple types of spatial queries on MapReduce through spatial partitioning, customizable spatial query engine RESQUE, implicit parallel spatial query execution on MapReduce, and effective methods for amending query results through handling boundary objects. Hadoop-GIS utilizes global partition indexing and customizable on demand local spatial indexing to achieve efficient query processing. Hadoop-GIS is integrated into Hive to support declarative spatial queries with an integrated architecture. Our experiments have demonstrated the high efficiency of Hadoop-GIS on query response and high scalability to run on commodity clusters. Our comparative experiments have showed that performance of Hadoop-GIS is on par with parallel SDBMS and outperforms SDBMS for compute-intensive queries. Hadoop-GIS is available as a set of library for processing spatial queries, and as an integrated software package in Hive.

The Computational Complexity, Parallel Scalability, and Performance of Atmospheric Data Assimilation Algorithms

NASA Technical Reports Server (NTRS)

Lyster, Peter M.; Guo, J.; Clune, T.; Larson, J. W.; Atlas, Robert (Technical Monitor)

2001-01-01

The computational complexity of algorithms for Four Dimensional Data Assimilation (4DDA) at NASA's Data Assimilation Office (DAO) is discussed. In 4DDA, observations are assimilated with the output of a dynamical model to generate best-estimates of the states of the system. It is thus a mapping problem, whereby scattered observations are converted into regular accurate maps of wind, temperature, moisture and other variables. The DAO is developing and using 4DDA algorithms that provide these datasets, or analyses, in support of Earth System Science research. Two large-scale algorithms are discussed. The first approach, the Goddard Earth Observing System Data Assimilation System (GEOS DAS), uses an atmospheric general circulation model (GCM) and an observation-space based analysis system, the Physical-space Statistical Analysis System (PSAS). GEOS DAS is very similar to global meteorological weather forecasting data assimilation systems, but is used at NASA for climate research. Systems of this size typically run at between 1 and 20 gigaflop/s. The second approach, the Kalman filter, uses a more consistent algorithm to determine the forecast error covariance matrix than does GEOS DAS. For atmospheric assimilation, the gridded dynamical fields typically have More than 10(exp 6) variables, therefore the full error covariance matrix may be in excess of a teraword. For the Kalman filter this problem can easily scale to petaflop/s proportions. We discuss the computational complexity of GEOS DAS and our implementation of the Kalman filter. We also discuss and quantify some of the technical issues and limitations in developing efficient, in terms of wall clock time, and scalable parallel implementations of the algorithms.

Hadoop-GIS: A High Performance Spatial Data Warehousing System over MapReduce

PubMed Central

Aji, Ablimit; Wang, Fusheng; Vo, Hoang; Lee, Rubao; Liu, Qiaoling; Zhang, Xiaodong; Saltz, Joel

2013-01-01

Support of high performance queries on large volumes of spatial data becomes increasingly important in many application domains, including geospatial problems in numerous fields, location based services, and emerging scientific applications that are increasingly data- and compute-intensive. The emergence of massive scale spatial data is due to the proliferation of cost effective and ubiquitous positioning technologies, development of high resolution imaging technologies, and contribution from a large number of community users. There are two major challenges for managing and querying massive spatial data to support spatial queries: the explosion of spatial data, and the high computational complexity of spatial queries. In this paper, we present Hadoop-GIS – a scalable and high performance spatial data warehousing system for running large scale spatial queries on Hadoop. Hadoop-GIS supports multiple types of spatial queries on MapReduce through spatial partitioning, customizable spatial query engine RESQUE, implicit parallel spatial query execution on MapReduce, and effective methods for amending query results through handling boundary objects. Hadoop-GIS utilizes global partition indexing and customizable on demand local spatial indexing to achieve efficient query processing. Hadoop-GIS is integrated into Hive to support declarative spatial queries with an integrated architecture. Our experiments have demonstrated the high efficiency of Hadoop-GIS on query response and high scalability to run on commodity clusters. Our comparative experiments have showed that performance of Hadoop-GIS is on par with parallel SDBMS and outperforms SDBMS for compute-intensive queries. Hadoop-GIS is available as a set of library for processing spatial queries, and as an integrated software package in Hive. PMID:24187650

CloudMC: a cloud computing application for Monte Carlo simulation.

PubMed

Miras, H; Jiménez, R; Miras, C; Gomà, C

2013-04-21

This work presents CloudMC, a cloud computing application-developed in Windows Azure®, the platform of the Microsoft® cloud-for the parallelization of Monte Carlo simulations in a dynamic virtual cluster. CloudMC is a web application designed to be independent of the Monte Carlo code in which the simulations are based-the simulations just need to be of the form: input files → executable → output files. To study the performance of CloudMC in Windows Azure®, Monte Carlo simulations with penelope were performed on different instance (virtual machine) sizes, and for different number of instances. The instance size was found to have no effect on the simulation runtime. It was also found that the decrease in time with the number of instances followed Amdahl's law, with a slight deviation due to the increase in the fraction of non-parallelizable time with increasing number of instances. A simulation that would have required 30 h of CPU on a single instance was completed in 48.6 min when executed on 64 instances in parallel (speedup of 37 ×). Furthermore, the use of cloud computing for parallel computing offers some advantages over conventional clusters: high accessibility, scalability and pay per usage. Therefore, it is strongly believed that cloud computing will play an important role in making Monte Carlo dose calculation a reality in future clinical practice.

Cpu/gpu Computing for AN Implicit Multi-Block Compressible Navier-Stokes Solver on Heterogeneous Platform

NASA Astrophysics Data System (ADS)

Deng, Liang; Bai, Hanli; Wang, Fang; Xu, Qingxin

2016-06-01

CPU/GPU computing allows scientists to tremendously accelerate their numerical codes. In this paper, we port and optimize a double precision alternating direction implicit (ADI) solver for three-dimensional compressible Navier-Stokes equations from our in-house Computational Fluid Dynamics (CFD) software on heterogeneous platform. First, we implement a full GPU version of the ADI solver to remove a lot of redundant data transfers between CPU and GPU, and then design two fine-grain schemes, namely “one-thread-one-point” and “one-thread-one-line”, to maximize the performance. Second, we present a dual-level parallelization scheme using the CPU/GPU collaborative model to exploit the computational resources of both multi-core CPUs and many-core GPUs within the heterogeneous platform. Finally, considering the fact that memory on a single node becomes inadequate when the simulation size grows, we present a tri-level hybrid programming pattern MPI-OpenMP-CUDA that merges fine-grain parallelism using OpenMP and CUDA threads with coarse-grain parallelism using MPI for inter-node communication. We also propose a strategy to overlap the computation with communication using the advanced features of CUDA and MPI programming. We obtain speedups of 6.0 for the ADI solver on one Tesla M2050 GPU in contrast to two Xeon X5670 CPUs. Scalability tests show that our implementation can offer significant performance improvement on heterogeneous platform.

Distributed Parallel Processing and Dynamic Load Balancing Techniques for Multidisciplinary High Speed Aircraft Design

NASA Technical Reports Server (NTRS)

Krasteva, Denitza T.

1998-01-01

Multidisciplinary design optimization (MDO) for large-scale engineering problems poses many challenges (e.g., the design of an efficient concurrent paradigm for global optimization based on disciplinary analyses, expensive computations over vast data sets, etc.) This work focuses on the application of distributed schemes for massively parallel architectures to MDO problems, as a tool for reducing computation time and solving larger problems. The specific problem considered here is configuration optimization of a high speed civil transport (HSCT), and the efficient parallelization of the embedded paradigm for reasonable design space identification. Two distributed dynamic load balancing techniques (random polling and global round robin with message combining) and two necessary termination detection schemes (global task count and token passing) were implemented and evaluated in terms of effectiveness and scalability to large problem sizes and a thousand processors. The effect of certain parameters on execution time was also inspected. Empirical results demonstrated stable performance and effectiveness for all schemes, and the parametric study showed that the selected algorithmic parameters have a negligible effect on performance.

A Domain-Decomposed Multilevel Method for Adaptively Refined Cartesian Grids with Embedded Boundaries

NASA Technical Reports Server (NTRS)

Aftosmis, M. J.; Berger, M. J.; Adomavicius, G.

2000-01-01

Preliminary verification and validation of an efficient Euler solver for adaptively refined Cartesian meshes with embedded boundaries is presented. The parallel, multilevel method makes use of a new on-the-fly parallel domain decomposition strategy based upon the use of space-filling curves, and automatically generates a sequence of coarse meshes for processing by the multigrid smoother. The coarse mesh generation algorithm produces grids which completely cover the computational domain at every level in the mesh hierarchy. A series of examples on realistically complex three-dimensional configurations demonstrate that this new coarsening algorithm reliably achieves mesh coarsening ratios in excess of 7 on adaptively refined meshes. Numerical investigations of the scheme's local truncation error demonstrate an achieved order of accuracy between 1.82 and 1.88. Convergence results for the multigrid scheme are presented for both subsonic and transonic test cases and demonstrate W-cycle multigrid convergence rates between 0.84 and 0.94. Preliminary parallel scalability tests on both simple wing and complex complete aircraft geometries shows a computational speedup of 52 on 64 processors using the run-time mesh partitioner.

Empirical study of parallel LRU simulation algorithms

NASA Technical Reports Server (NTRS)

Carr, Eric; Nicol, David M.

1994-01-01

This paper reports on the performance of five parallel algorithms for simulating a fully associative cache operating under the LRU (Least-Recently-Used) replacement policy. Three of the algorithms are SIMD, and are implemented on the MasPar MP-2 architecture. Two other algorithms are parallelizations of an efficient serial algorithm on the Intel Paragon. One SIMD algorithm is quite simple, but its cost is linear in the cache size. The two other SIMD algorithm are more complex, but have costs that are independent on the cache size. Both the second and third SIMD algorithms compute all stack distances; the second SIMD algorithm is completely general, whereas the third SIMD algorithm presumes and takes advantage of bounds on the range of reference tags. Both MIMD algorithm implemented on the Paragon are general and compute all stack distances; they differ in one step that may affect their respective scalability. We assess the strengths and weaknesses of these algorithms as a function of problem size and characteristics, and compare their performance on traces derived from execution of three SPEC benchmark programs.

Adaptation of a Multi-Block Structured Solver for Effective Use in a Hybrid CPU/GPU Massively Parallel Environment

NASA Astrophysics Data System (ADS)

Gutzwiller, David; Gontier, Mathieu; Demeulenaere, Alain

2014-11-01

Multi-Block structured solvers hold many advantages over their unstructured counterparts, such as a smaller memory footprint and efficient serial performance. Historically, multi-block structured solvers have not been easily adapted for use in a High Performance Computing (HPC) environment, and the recent trend towards hybrid GPU/CPU architectures has further complicated the situation. This paper will elaborate on developments and innovations applied to the NUMECA FINE/Turbo solver that have allowed near-linear scalability with real-world problems on over 250 hybrid GPU/GPU cluster nodes. Discussion will focus on the implementation of virtual partitioning and load balancing algorithms using a novel meta-block concept. This implementation is transparent to the user, allowing all pre- and post-processing steps to be performed using a simple, unpartitioned grid topology. Additional discussion will elaborate on developments that have improved parallel performance, including fully parallel I/O with the ADIOS API and the GPU porting of the computationally heavy CPUBooster convergence acceleration module. Head of HPC and Release Management, Numeca International.

Improving the scalability of hyperspectral imaging applications on heterogeneous platforms using adaptive run-time data compression

NASA Astrophysics Data System (ADS)

Plaza, Antonio; Plaza, Javier; Paz, Abel

2010-10-01

Latest generation remote sensing instruments (called hyperspectral imagers) are now able to generate hundreds of images, corresponding to different wavelength channels, for the same area on the surface of the Earth. In previous work, we have reported that the scalability of parallel processing algorithms dealing with these high-dimensional data volumes is affected by the amount of data to be exchanged through the communication network of the system. However, large messages are common in hyperspectral imaging applications since processing algorithms are pixel-based, and each pixel vector to be exchanged through the communication network is made up of hundreds of spectral values. Thus, decreasing the amount of data to be exchanged could improve the scalability and parallel performance. In this paper, we propose a new framework based on intelligent utilization of wavelet-based data compression techniques for improving the scalability of a standard hyperspectral image processing chain on heterogeneous networks of workstations. This type of parallel platform is quickly becoming a standard in hyperspectral image processing due to the distributed nature of collected hyperspectral data as well as its flexibility and low cost. Our experimental results indicate that adaptive lossy compression can lead to improvements in the scalability of the hyperspectral processing chain without sacrificing analysis accuracy, even at sub-pixel precision levels.

FWT2D: A massively parallel program for frequency-domain full-waveform tomography of wide-aperture seismic data—Part 1: Algorithm

NASA Astrophysics Data System (ADS)

Sourbier, Florent; Operto, Stéphane; Virieux, Jean; Amestoy, Patrick; L'Excellent, Jean-Yves

2009-03-01

This is the first paper in a two-part series that describes a massively parallel code that performs 2D frequency-domain full-waveform inversion of wide-aperture seismic data for imaging complex structures. Full-waveform inversion methods, namely quantitative seismic imaging methods based on the resolution of the full wave equation, are computationally expensive. Therefore, designing efficient algorithms which take advantage of parallel computing facilities is critical for the appraisal of these approaches when applied to representative case studies and for further improvements. Full-waveform modelling requires the resolution of a large sparse system of linear equations which is performed with the massively parallel direct solver MUMPS for efficient multiple-shot simulations. Efficiency of the multiple-shot solution phase (forward/backward substitutions) is improved by using the BLAS3 library. The inverse problem relies on a classic local optimization approach implemented with a gradient method. The direct solver returns the multiple-shot wavefield solutions distributed over the processors according to a domain decomposition driven by the distribution of the LU factors. The domain decomposition of the wavefield solutions is used to compute in parallel the gradient of the objective function and the diagonal Hessian, this latter providing a suitable scaling of the gradient. The algorithm allows one to test different strategies for multiscale frequency inversion ranging from successive mono-frequency inversion to simultaneous multifrequency inversion. These different inversion strategies will be illustrated in the following companion paper. The parallel efficiency and the scalability of the code will also be quantified.

A scalable approach to solving dense linear algebra problems on hybrid CPU-GPU systems

DOE PAGES

Song, Fengguang; Dongarra, Jack

2014-10-01

Aiming to fully exploit the computing power of all CPUs and all graphics processing units (GPUs) on hybrid CPU-GPU systems to solve dense linear algebra problems, in this paper we design a class of heterogeneous tile algorithms to maximize the degree of parallelism, to minimize the communication volume, and to accommodate the heterogeneity between CPUs and GPUs. The new heterogeneous tile algorithms are executed upon our decentralized dynamic scheduling runtime system, which schedules a task graph dynamically and transfers data between compute nodes automatically. The runtime system uses a new distributed task assignment protocol to solve data dependencies between tasksmore » without any coordination between processing units. By overlapping computation and communication through dynamic scheduling, we are able to attain scalable performance for the double-precision Cholesky factorization and QR factorization. Finally, our approach demonstrates a performance comparable to Intel MKL on shared-memory multicore systems and better performance than both vendor (e.g., Intel MKL) and open source libraries (e.g., StarPU) in the following three environments: heterogeneous clusters with GPUs, conventional clusters without GPUs, and shared-memory systems with multiple GPUs.« less

A scalable approach to solving dense linear algebra problems on hybrid CPU-GPU systems

DOE Office of Scientific and Technical Information (OSTI.GOV)

Song, Fengguang; Dongarra, Jack

Aiming to fully exploit the computing power of all CPUs and all graphics processing units (GPUs) on hybrid CPU-GPU systems to solve dense linear algebra problems, in this paper we design a class of heterogeneous tile algorithms to maximize the degree of parallelism, to minimize the communication volume, and to accommodate the heterogeneity between CPUs and GPUs. The new heterogeneous tile algorithms are executed upon our decentralized dynamic scheduling runtime system, which schedules a task graph dynamically and transfers data between compute nodes automatically. The runtime system uses a new distributed task assignment protocol to solve data dependencies between tasksmore » without any coordination between processing units. By overlapping computation and communication through dynamic scheduling, we are able to attain scalable performance for the double-precision Cholesky factorization and QR factorization. Finally, our approach demonstrates a performance comparable to Intel MKL on shared-memory multicore systems and better performance than both vendor (e.g., Intel MKL) and open source libraries (e.g., StarPU) in the following three environments: heterogeneous clusters with GPUs, conventional clusters without GPUs, and shared-memory systems with multiple GPUs.« less

Fast MPEG-CDVS Encoder With GPU-CPU Hybrid Computing

NASA Astrophysics Data System (ADS)

Duan, Ling-Yu; Sun, Wei; Zhang, Xinfeng; Wang, Shiqi; Chen, Jie; Yin, Jianxiong; See, Simon; Huang, Tiejun; Kot, Alex C.; Gao, Wen

2018-05-01

The compact descriptors for visual search (CDVS) standard from ISO/IEC moving pictures experts group (MPEG) has succeeded in enabling the interoperability for efficient and effective image retrieval by standardizing the bitstream syntax of compact feature descriptors. However, the intensive computation of CDVS encoder unfortunately hinders its widely deployment in industry for large-scale visual search. In this paper, we revisit the merits of low complexity design of CDVS core techniques and present a very fast CDVS encoder by leveraging the massive parallel execution resources of GPU. We elegantly shift the computation-intensive and parallel-friendly modules to the state-of-the-arts GPU platforms, in which the thread block allocation and the memory access are jointly optimized to eliminate performance loss. In addition, those operations with heavy data dependence are allocated to CPU to resolve the extra but non-necessary computation burden for GPU. Furthermore, we have demonstrated the proposed fast CDVS encoder can work well with those convolution neural network approaches which has harmoniously leveraged the advantages of GPU platforms, and yielded significant performance improvements. Comprehensive experimental results over benchmarks are evaluated, which has shown that the fast CDVS encoder using GPU-CPU hybrid computing is promising for scalable visual search.

Iterative load-balancing method with multigrid level relaxation for particle simulation with short-range interactions

NASA Astrophysics Data System (ADS)

Furuichi, Mikito; Nishiura, Daisuke

2017-10-01

We developed dynamic load-balancing algorithms for Particle Simulation Methods (PSM) involving short-range interactions, such as Smoothed Particle Hydrodynamics (SPH), Moving Particle Semi-implicit method (MPS), and Discrete Element method (DEM). These are needed to handle billions of particles modeled in large distributed-memory computer systems. Our method utilizes flexible orthogonal domain decomposition, allowing the sub-domain boundaries in the column to be different for each row. The imbalances in the execution time between parallel logical processes are treated as a nonlinear residual. Load-balancing is achieved by minimizing the residual within the framework of an iterative nonlinear solver, combined with a multigrid technique in the local smoother. Our iterative method is suitable for adjusting the sub-domain frequently by monitoring the performance of each computational process because it is computationally cheaper in terms of communication and memory costs than non-iterative methods. Numerical tests demonstrated the ability of our approach to handle workload imbalances arising from a non-uniform particle distribution, differences in particle types, or heterogeneous computer architecture which was difficult with previously proposed methods. We analyzed the parallel efficiency and scalability of our method using Earth simulator and K-computer supercomputer systems.

Parallelisation study of a three-dimensional environmental flow model

NASA Astrophysics Data System (ADS)

O'Donncha, Fearghal; Ragnoli, Emanuele; Suits, Frank

2014-03-01

There are many simulation codes in the geosciences that are serial and cannot take advantage of the parallel computational resources commonly available today. One model important for our work in coastal ocean current modelling is EFDC, a Fortran 77 code configured for optimal deployment on vector computers. In order to take advantage of our cache-based, blade computing system we restructured EFDC from serial to parallel, thereby allowing us to run existing models more quickly, and to simulate larger and more detailed models that were previously impractical. Since the source code for EFDC is extensive and involves detailed computation, it is important to do such a port in a manner that limits changes to the files, while achieving the desired speedup. We describe a parallelisation strategy involving surgical changes to the source files to minimise error-prone alteration of the underlying computations, while allowing load-balanced domain decomposition for efficient execution on a commodity cluster. The use of conjugate gradient posed particular challenges due to implicit non-local communication posing a hindrance to standard domain partitioning schemes; a number of techniques are discussed to address this in a feasible, computationally efficient manner. The parallel implementation demonstrates good scalability in combination with a novel domain partitioning scheme that specifically handles mixed water/land regions commonly found in coastal simulations. The approach presented here represents a practical methodology to rejuvenate legacy code on a commodity blade cluster with reasonable effort; our solution has direct application to other similar codes in the geosciences.

A Lightweight Remote Parallel Visualization Platform for Interactive Massive Time-varying Climate Data Analysis

NASA Astrophysics Data System (ADS)

Li, J.; Zhang, T.; Huang, Q.; Liu, Q.

2014-12-01

Today's climate datasets are featured with large volume, high degree of spatiotemporal complexity and evolving fast overtime. As visualizing large volume distributed climate datasets is computationally intensive, traditional desktop based visualization applications fail to handle the computational intensity. Recently, scientists have developed remote visualization techniques to address the computational issue. Remote visualization techniques usually leverage server-side parallel computing capabilities to perform visualization tasks and deliver visualization results to clients through network. In this research, we aim to build a remote parallel visualization platform for visualizing and analyzing massive climate data. Our visualization platform was built based on Paraview, which is one of the most popular open source remote visualization and analysis applications. To further enhance the scalability and stability of the platform, we have employed cloud computing techniques to support the deployment of the platform. In this platform, all climate datasets are regular grid data which are stored in NetCDF format. Three types of data access methods are supported in the platform: accessing remote datasets provided by OpenDAP servers, accessing datasets hosted on the web visualization server and accessing local datasets. Despite different data access methods, all visualization tasks are completed at the server side to reduce the workload of clients. As a proof of concept, we have implemented a set of scientific visualization methods to show the feasibility of the platform. Preliminary results indicate that the framework can address the computation limitation of desktop based visualization applications.

Survey of MapReduce frame operation in bioinformatics.

PubMed

Zou, Quan; Li, Xu-Bin; Jiang, Wen-Rui; Lin, Zi-Yu; Li, Gui-Lin; Chen, Ke

2014-07-01

Bioinformatics is challenged by the fact that traditional analysis tools have difficulty in processing large-scale data from high-throughput sequencing. The open source Apache Hadoop project, which adopts the MapReduce framework and a distributed file system, has recently given bioinformatics researchers an opportunity to achieve scalable, efficient and reliable computing performance on Linux clusters and on cloud computing services. In this article, we present MapReduce frame-based applications that can be employed in the next-generation sequencing and other biological domains. In addition, we discuss the challenges faced by this field as well as the future works on parallel computing in bioinformatics. © The Author 2013. Published by Oxford University Press. For Permissions, please email: journals.permissions@oup.com.

Parallel computing of physical maps--a comparative study in SIMD and MIMD parallelism.

PubMed

Bhandarkar, S M; Chirravuri, S; Arnold, J

1996-01-01

Ordering clones from a genomic library into physical maps of whole chromosomes presents a central computational problem in genetics. Chromosome reconstruction via clone ordering is usually isomorphic to the NP-complete Optimal Linear Arrangement problem. Parallel SIMD and MIMD algorithms for simulated annealing based on Markov chain distribution are proposed and applied to the problem of chromosome reconstruction via clone ordering. Perturbation methods and problem-specific annealing heuristics are proposed and described. The SIMD algorithms are implemented on a 2048 processor MasPar MP-2 system which is an SIMD 2-D toroidal mesh architecture whereas the MIMD algorithms are implemented on an 8 processor Intel iPSC/860 which is an MIMD hypercube architecture. A comparative analysis of the various SIMD and MIMD algorithms is presented in which the convergence, speedup, and scalability characteristics of the various algorithms are analyzed and discussed. On a fine-grained, massively parallel SIMD architecture with a low synchronization overhead such as the MasPar MP-2, a parallel simulated annealing algorithm based on multiple periodically interacting searches performs the best. For a coarse-grained MIMD architecture with high synchronization overhead such as the Intel iPSC/860, a parallel simulated annealing algorithm based on multiple independent searches yields the best results. In either case, distribution of clonal data across multiple processors is shown to exacerbate the tendency of the parallel simulated annealing algorithm to get trapped in a local optimum.

birgHPC: creating instant computing clusters for bioinformatics and molecular dynamics.

PubMed

Chew, Teong Han; Joyce-Tan, Kwee Hong; Akma, Farizuwana; Shamsir, Mohd Shahir

2011-05-01

birgHPC, a bootable Linux Live CD has been developed to create high-performance clusters for bioinformatics and molecular dynamics studies using any Local Area Network (LAN)-networked computers. birgHPC features automated hardware and slots detection as well as provides a simple job submission interface. The latest versions of GROMACS, NAMD, mpiBLAST and ClustalW-MPI can be run in parallel by simply booting the birgHPC CD or flash drive from the head node, which immediately positions the rest of the PCs on the network as computing nodes. Thus, a temporary, affordable, scalable and high-performance computing environment can be built by non-computing-based researchers using low-cost commodity hardware. The birgHPC Live CD and relevant user guide are available for free at http://birg1.fbb.utm.my/birghpc.

Tinker-HP: a massively parallel molecular dynamics package for multiscale simulations of large complex systems with advanced point dipole polarizable force fields† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c7sc04531j

PubMed Central

Lagardère, Louis; Jolly, Luc-Henri; Lipparini, Filippo; Aviat, Félix; Stamm, Benjamin; Jing, Zhifeng F.; Harger, Matthew; Torabifard, Hedieh; Cisneros, G. Andrés; Schnieders, Michael J.; Gresh, Nohad; Maday, Yvon; Ren, Pengyu Y.; Ponder, Jay W.

2017-01-01

We present Tinker-HP, a massively MPI parallel package dedicated to classical molecular dynamics (MD) and to multiscale simulations, using advanced polarizable force fields (PFF) encompassing distributed multipoles electrostatics. Tinker-HP is an evolution of the popular Tinker package code that conserves its simplicity of use and its reference double precision implementation for CPUs. Grounded on interdisciplinary efforts with applied mathematics, Tinker-HP allows for long polarizable MD simulations on large systems up to millions of atoms. We detail in the paper the newly developed extension of massively parallel 3D spatial decomposition to point dipole polarizable models as well as their coupling to efficient Krylov iterative and non-iterative polarization solvers. The design of the code allows the use of various computer systems ranging from laboratory workstations to modern petascale supercomputers with thousands of cores. Tinker-HP proposes therefore the first high-performance scalable CPU computing environment for the development of next generation point dipole PFFs and for production simulations. Strategies linking Tinker-HP to Quantum Mechanics (QM) in the framework of multiscale polarizable self-consistent QM/MD simulations are also provided. The possibilities, performances and scalability of the software are demonstrated via benchmarks calculations using the polarizable AMOEBA force field on systems ranging from large water boxes of increasing size and ionic liquids to (very) large biosystems encompassing several proteins as well as the complete satellite tobacco mosaic virus and ribosome structures. For small systems, Tinker-HP appears to be competitive with the Tinker-OpenMM GPU implementation of Tinker. As the system size grows, Tinker-HP remains operational thanks to its access to distributed memory and takes advantage of its new algorithmic enabling for stable long timescale polarizable simulations. Overall, a several thousand-fold acceleration over a single-core computation is observed for the largest systems. The extension of the present CPU implementation of Tinker-HP to other computational platforms is discussed. PMID:29732110

Disparity : scalable anomaly detection for clusters.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Desai, N.; Bradshaw, R.; Lusk, E.

2008-01-01

In this paper, we describe disparity, a tool that does parallel, scalable anomaly detection for clusters. Disparity uses basic statistical methods and scalable reduction operations to perform data reduction on client nodes and uses these results to locate node anomalies. We discuss the implementation of disparity and present results of its use on a SiCortex SC5832 system.

P43-S Computational Biology Applications Suite for High-Performance Computing (BioHPC.net)

PubMed Central

Pillardy, J.

2007-01-01

One of the challenges of high-performance computing (HPC) is user accessibility. At the Cornell University Computational Biology Service Unit, which is also a Microsoft HPC institute, we have developed a computational biology application suite that allows researchers from biological laboratories to submit their jobs to the parallel cluster through an easy-to-use Web interface. Through this system, we are providing users with popular bioinformatics tools including BLAST, HMMER, InterproScan, and MrBayes. The system is flexible and can be easily customized to include other software. It is also scalable; the installation on our servers currently processes approximately 8500 job submissions per year, many of them requiring massively parallel computations. It also has a built-in user management system, which can limit software and/or database access to specified users. TAIR, the major database of the plant model organism Arabidopsis, and SGN, the international tomato genome database, are both using our system for storage and data analysis. The system consists of a Web server running the interface (ASP.NET C#), Microsoft SQL server (ADO.NET), compute cluster running Microsoft Windows, ftp server, and file server. Users can interact with their jobs and data via a Web browser, ftp, or e-mail. The interface is accessible at http://cbsuapps.tc.cornell.edu/.

Approaching the exa-scale: a real-world evaluation of rendering extremely large data sets

DOE Office of Scientific and Technical Information (OSTI.GOV)

Patchett, John M; Ahrens, James P; Lo, Li - Ta

2010-10-15

Extremely large scale analysis is becoming increasingly important as supercomputers and their simulations move from petascale to exascale. The lack of dedicated hardware acceleration for rendering on today's supercomputing platforms motivates our detailed evaluation of the possibility of interactive rendering on the supercomputer. In order to facilitate our understanding of rendering on the supercomputing platform, we focus on scalability of rendering algorithms and architecture envisioned for exascale datasets. To understand tradeoffs for dealing with extremely large datasets, we compare three different rendering algorithms for large polygonal data: software based ray tracing, software based rasterization and hardware accelerated rasterization. We presentmore » a case study of strong and weak scaling of rendering extremely large data on both GPU and CPU based parallel supercomputers using Para View, a parallel visualization tool. Wc use three different data sets: two synthetic and one from a scientific application. At an extreme scale, algorithmic rendering choices make a difference and should be considered while approaching exascale computing, visualization, and analysis. We find software based ray-tracing offers a viable approach for scalable rendering of the projected future massive data sizes.« less

Optimization of atmospheric transport models on HPC platforms

NASA Astrophysics Data System (ADS)

de la Cruz, Raúl; Folch, Arnau; Farré, Pau; Cabezas, Javier; Navarro, Nacho; Cela, José María

2016-12-01

The performance and scalability of atmospheric transport models on high performance computing environments is often far from optimal for multiple reasons including, for example, sequential input and output, synchronous communications, work unbalance, memory access latency or lack of task overlapping. We investigate how different software optimizations and porting to non general-purpose hardware architectures improve code scalability and execution times considering, as an example, the FALL3D volcanic ash transport model. To this purpose, we implement the FALL3D model equations in the WARIS framework, a software designed from scratch to solve in a parallel and efficient way different geoscience problems on a wide variety of architectures. In addition, we consider further improvements in WARIS such as hybrid MPI-OMP parallelization, spatial blocking, auto-tuning and thread affinity. Considering all these aspects together, the FALL3D execution times for a realistic test case running on general-purpose cluster architectures (Intel Sandy Bridge) decrease by a factor between 7 and 40 depending on the grid resolution. Finally, we port the application to Intel Xeon Phi (MIC) and NVIDIA GPUs (CUDA) accelerator-based architectures and compare performance, cost and power consumption on all the architectures. Implications on time-constrained operational model configurations are discussed.

Parallel Programming Strategies for Irregular Adaptive Applications

NASA Technical Reports Server (NTRS)

Biswas, Rupak; Biegel, Bryan (Technical Monitor)

2001-01-01

Achieving scalable performance for dynamic irregular applications is eminently challenging. Traditional message-passing approaches have been making steady progress towards this goal; however, they suffer from complex implementation requirements. The use of a global address space greatly simplifies the programming task, but can degrade the performance for such computations. In this work, we examine two typical irregular adaptive applications, Dynamic Remeshing and N-Body, under competing programming methodologies and across various parallel architectures. The Dynamic Remeshing application simulates flow over an airfoil, and refines localized regions of the underlying unstructured mesh. The N-Body experiment models two neighboring Plummer galaxies that are about to undergo a merger. Both problems demonstrate dramatic changes in processor workloads and interprocessor communication with time; thus, dynamic load balancing is a required component.

SU-E-T-531: Performance Evaluation of Multithreaded Geant4 for Proton Therapy Dose Calculations in a High Performance Computing Facility

DOE Office of Scientific and Technical Information (OSTI.GOV)

Shin, J; Coss, D; McMurry, J

Purpose: To evaluate the efficiency of multithreaded Geant4 (Geant4-MT, version 10.0) for proton Monte Carlo dose calculations using a high performance computing facility. Methods: Geant4-MT was used to calculate 3D dose distributions in 1×1×1 mm3 voxels in a water phantom and patient's head with a 150 MeV proton beam covering approximately 5×5 cm2 in the water phantom. Three timestamps were measured on the fly to separately analyze the required time for initialization (which cannot be parallelized), processing time of individual threads, and completion time. Scalability of averaged processing time per thread was calculated as a function of thread number (1,more » 100, 150, and 200) for both 1M and 50 M histories. The total memory usage was recorded. Results: Simulations with 50 M histories were fastest with 100 threads, taking approximately 1.3 hours and 6 hours for the water phantom and the CT data, respectively with better than 1.0 % statistical uncertainty. The calculations show 1/N scalability in the event loops for both cases. The gains from parallel calculations started to decrease with 150 threads. The memory usage increases linearly with number of threads. No critical failures were observed during the simulations. Conclusion: Multithreading in Geant4-MT decreased simulation time in proton dose distribution calculations by a factor of 64 and 54 at a near optimal 100 threads for water phantom and patient's data respectively. Further simulations will be done to determine the efficiency at the optimal thread number. Considering the trend of computer architecture development, utilizing Geant4-MT for radiotherapy simulations is an excellent cost-effective alternative for a distributed batch queuing system. However, because the scalability depends highly on simulation details, i.e., the ratio of the processing time of one event versus waiting time to access for the shared event queue, a performance evaluation as described is recommended.« less

The Parallel System for Integrating Impact Models and Sectors (pSIMS)

NASA Technical Reports Server (NTRS)

Elliott, Joshua; Kelly, David; Chryssanthacopoulos, James; Glotter, Michael; Jhunjhnuwala, Kanika; Best, Neil; Wilde, Michael; Foster, Ian

2014-01-01

We present a framework for massively parallel climate impact simulations: the parallel System for Integrating Impact Models and Sectors (pSIMS). This framework comprises a) tools for ingesting and converting large amounts of data to a versatile datatype based on a common geospatial grid; b) tools for translating this datatype into custom formats for site-based models; c) a scalable parallel framework for performing large ensemble simulations, using any one of a number of different impacts models, on clusters, supercomputers, distributed grids, or clouds; d) tools and data standards for reformatting outputs to common datatypes for analysis and visualization; and e) methodologies for aggregating these datatypes to arbitrary spatial scales such as administrative and environmental demarcations. By automating many time-consuming and error-prone aspects of large-scale climate impacts studies, pSIMS accelerates computational research, encourages model intercomparison, and enhances reproducibility of simulation results. We present the pSIMS design and use example assessments to demonstrate its multi-model, multi-scale, and multi-sector versatility.

A general parallel sparse-blocked matrix multiply for linear scaling SCF theory

NASA Astrophysics Data System (ADS)

Challacombe, Matt

2000-06-01

A general approach to the parallel sparse-blocked matrix-matrix multiply is developed in the context of linear scaling self-consistent-field (SCF) theory. The data-parallel message passing method uses non-blocking communication to overlap computation and communication. The space filling curve heuristic is used to achieve data locality for sparse matrix elements that decay with “separation”. Load balance is achieved by solving the bin packing problem for blocks with variable size.With this new method as the kernel, parallel performance of the simplified density matrix minimization (SDMM) for solution of the SCF equations is investigated for RHF/6-31G ∗∗ water clusters and RHF/3-21G estane globules. Sustained rates above 5.7 GFLOPS for the SDMM have been achieved for (H 2 O) 200 with 95 Origin 2000 processors. Scalability is found to be limited by load imbalance, which increases with decreasing granularity, due primarily to the inhomogeneous distribution of variable block sizes.

Full Parallel Implementation of an All-Electron Four-Component Dirac-Kohn-Sham Program.

PubMed

Rampino, Sergio; Belpassi, Leonardo; Tarantelli, Francesco; Storchi, Loriano

2014-09-09

A full distributed-memory implementation of the Dirac-Kohn-Sham (DKS) module of the program BERTHA (Belpassi et al., Phys. Chem. Chem. Phys. 2011, 13, 12368-12394) is presented, where the self-consistent field (SCF) procedure is replicated on all the parallel processes, each process working on subsets of the global matrices. The key feature of the implementation is an efficient procedure for switching between two matrix distribution schemes, one (integral-driven) optimal for the parallel computation of the matrix elements and another (block-cyclic) optimal for the parallel linear algebra operations. This approach, making both CPU-time and memory scalable with the number of processors used, virtually overcomes at once both time and memory barriers associated with DKS calculations. Performance, portability, and numerical stability of the code are illustrated on the basis of test calculations on three gold clusters of increasing size, an organometallic compound, and a perovskite model. The calculations are performed on a Beowulf and a BlueGene/Q system.

Implementation, capabilities, and benchmarking of Shift, a massively parallel Monte Carlo radiation transport code

DOE PAGES

Pandya, Tara M.; Johnson, Seth R.; Evans, Thomas M.; ...

2015-12-21

This paper discusses the implementation, capabilities, and validation of Shift, a massively parallel Monte Carlo radiation transport package developed and maintained at Oak Ridge National Laboratory. It has been developed to scale well from laptop to small computing clusters to advanced supercomputers. Special features of Shift include hybrid capabilities for variance reduction such as CADIS and FW-CADIS, and advanced parallel decomposition and tally methods optimized for scalability on supercomputing architectures. Shift has been validated and verified against various reactor physics benchmarks and compares well to other state-of-the-art Monte Carlo radiation transport codes such as MCNP5, CE KENO-VI, and OpenMC. Somemore » specific benchmarks used for verification and validation include the CASL VERA criticality test suite and several Westinghouse AP1000 ® problems. These benchmark and scaling studies show promising results.« less

Myria: Scalable Analytics as a Service

NASA Astrophysics Data System (ADS)

Howe, B.; Halperin, D.; Whitaker, A.

2014-12-01

At the UW eScience Institute, we're working to empower non-experts, especially in the sciences, to write and use data-parallel algorithms. To this end, we are building Myria, a web-based platform for scalable analytics and data-parallel programming. Myria's internal model of computation is the relational algebra extended with iteration, such that every program is inherently data-parallel, just as every query in a database is inherently data-parallel. But unlike databases, iteration is a first class concept, allowing us to express machine learning tasks, graph traversal tasks, and more. Programs can be expressed in a number of languages and can be executed on a number of execution environments, but we emphasize a particular language called MyriaL that supports both imperative and declarative styles and a particular execution engine called MyriaX that uses an in-memory column-oriented representation and asynchronous iteration. We deliver Myria over the web as a service, providing an editor, performance analysis tools, and catalog browsing features in a single environment. We find that this web-based "delivery vector" is critical in reaching non-experts: they are insulated from irrelevant effort technical work associated with installation, configuration, and resource management. The MyriaX backend, one of several execution runtimes we support, is a main-memory, column-oriented, RDBMS-on-the-worker system that supports cyclic data flows as a first-class citizen and has been shown to outperform competitive systems on 100-machine cluster sizes. I will describe the Myria system, give a demo, and present some new results in large-scale oceanographic microbiology.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Slattery, Stuart R.

In this study we analyze and extend mesh-free algorithms for three-dimensional data transfer problems in partitioned multiphysics simulations. We first provide a direct comparison between a mesh-based weighted residual method using the common-refinement scheme and two mesh-free algorithms leveraging compactly supported radial basis functions: one using a spline interpolation and one using a moving least square reconstruction. Through the comparison we assess both the conservation and accuracy of the data transfer obtained from each of the methods. We do so for a varying set of geometries with and without curvature and sharp features and for functions with and without smoothnessmore » and with varying gradients. Our results show that the mesh-based and mesh-free algorithms are complementary with cases where each was demonstrated to perform better than the other. We then focus on the mesh-free methods by developing a set of algorithms to parallelize them based on sparse linear algebra techniques. This includes a discussion of fast parallel radius searching in point clouds and restructuring the interpolation algorithms to leverage data structures and linear algebra services designed for large distributed computing environments. The scalability of our new algorithms is demonstrated on a leadership class computing facility using a set of basic scaling studies. Finally, these scaling studies show that for problems with reasonable load balance, our new algorithms for both spline interpolation and moving least square reconstruction demonstrate both strong and weak scalability using more than 100,000 MPI processes with billions of degrees of freedom in the data transfer operation.« less

paraGSEA: a scalable approach for large-scale gene expression profiling

PubMed Central

Peng, Shaoliang; Yang, Shunyun

2017-01-01

Abstract More studies have been conducted using gene expression similarity to identify functional connections among genes, diseases and drugs. Gene Set Enrichment Analysis (GSEA) is a powerful analytical method for interpreting gene expression data. However, due to its enormous computational overhead in the estimation of significance level step and multiple hypothesis testing step, the computation scalability and efficiency are poor on large-scale datasets. We proposed paraGSEA for efficient large-scale transcriptome data analysis. By optimization, the overall time complexity of paraGSEA is reduced from O(mn) to O(m+n), where m is the length of the gene sets and n is the length of the gene expression profiles, which contributes more than 100-fold increase in performance compared with other popular GSEA implementations such as GSEA-P, SAM-GS and GSEA2. By further parallelization, a near-linear speed-up is gained on both workstations and clusters in an efficient manner with high scalability and performance on large-scale datasets. The analysis time of whole LINCS phase I dataset (GSE92742) was reduced to nearly half hour on a 1000 node cluster on Tianhe-2, or within 120 hours on a 96-core workstation. The source code of paraGSEA is licensed under the GPLv3 and available at http://github.com/ysycloud/paraGSEA. PMID:28973463

Heterogeneous scalable framework for multiphase flows

DOE Office of Scientific and Technical Information (OSTI.GOV)

Morris, Karla Vanessa

2013-09-01

Two categories of challenges confront the developer of computational spray models: those related to the computation and those related to the physics. Regarding the computation, the trend towards heterogeneous, multi- and many-core platforms will require considerable re-engineering of codes written for the current supercomputing platforms. Regarding the physics, accurate methods for transferring mass, momentum and energy from the dispersed phase onto the carrier fluid grid have so far eluded modelers. Significant challenges also lie at the intersection between these two categories. To be competitive, any physics model must be expressible in a parallel algorithm that performs well on evolving computermore » platforms. This work created an application based on a software architecture where the physics and software concerns are separated in a way that adds flexibility to both. The develop spray-tracking package includes an application programming interface (API) that abstracts away the platform-dependent parallelization concerns, enabling the scientific programmer to write serial code that the API resolves into parallel processes and threads of execution. The project also developed the infrastructure required to provide similar APIs to other application. The API allow object-oriented Fortran applications direct interaction with Trilinos to support memory management of distributed objects in central processing units (CPU) and graphic processing units (GPU) nodes for applications using C++.« less

A Comparison of PETSC Library and HPF Implementations of an Archetypal PDE Computation

NASA Technical Reports Server (NTRS)

Hayder, M. Ehtesham; Keyes, David E.; Mehrotra, Piyush

1997-01-01

Two paradigms for distributed-memory parallel computation that free the application programmer from the details of message passing are compared for an archetypal structured scientific computation a nonlinear, structured-grid partial differential equation boundary value problem using the same algorithm on the same hardware. Both paradigms, parallel libraries represented by Argonne's PETSC, and parallel languages represented by the Portland Group's HPF, are found to be easy to use for this problem class, and both are reasonably effective in exploiting concurrency after a short learning curve. The level of involvement required by the application programmer under either paradigm includes specification of the data partitioning (corresponding to a geometrically simple decomposition of the domain of the PDE). Programming in SPAM style for the PETSC library requires writing the routines that discretize the PDE and its Jacobian, managing subdomain-to-processor mappings (affine global- to-local index mappings), and interfacing to library solver routines. Programming for HPF requires a complete sequential implementation of the same algorithm, introducing concurrency through subdomain blocking (an effort similar to the index mapping), and modest experimentation with rewriting loops to elucidate to the compiler the latent concurrency. Correctness and scalability are cross-validated on up to 32 nodes of an IBM SP2.

Approximate kernel competitive learning.

PubMed

Wu, Jian-Sheng; Zheng, Wei-Shi; Lai, Jian-Huang

2015-03-01

Kernel competitive learning has been successfully used to achieve robust clustering. However, kernel competitive learning (KCL) is not scalable for large scale data processing, because (1) it has to calculate and store the full kernel matrix that is too large to be calculated and kept in the memory and (2) it cannot be computed in parallel. In this paper we develop a framework of approximate kernel competitive learning for processing large scale dataset. The proposed framework consists of two parts. First, it derives an approximate kernel competitive learning (AKCL), which learns kernel competitive learning in a subspace via sampling. We provide solid theoretical analysis on why the proposed approximation modelling would work for kernel competitive learning, and furthermore, we show that the computational complexity of AKCL is largely reduced. Second, we propose a pseudo-parallelled approximate kernel competitive learning (PAKCL) based on a set-based kernel competitive learning strategy, which overcomes the obstacle of using parallel programming in kernel competitive learning and significantly accelerates the approximate kernel competitive learning for large scale clustering. The empirical evaluation on publicly available datasets shows that the proposed AKCL and PAKCL can perform comparably as KCL, with a large reduction on computational cost. Also, the proposed methods achieve more effective clustering performance in terms of clustering precision against related approximate clustering approaches. Copyright © 2014 Elsevier Ltd. All rights reserved.

Enabling High-performance Interactive Geoscience Data Analysis Through Data Placement and Movement Optimization

NASA Astrophysics Data System (ADS)

Zhu, F.; Yu, H.; Rilee, M. L.; Kuo, K. S.; Yu, L.; Pan, Y.; Jiang, H.

2017-12-01

Since the establishment of data archive centers and the standardization of file formats, scientists are required to search metadata catalogs for data needed and download the data files to their local machines to carry out data analysis. This approach has facilitated data discovery and access for decades, but it inevitably leads to data transfer from data archive centers to scientists' computers through low-bandwidth Internet connections. Data transfer becomes a major performance bottleneck in such an approach. Combined with generally constrained local compute/storage resources, they limit the extent of scientists' studies and deprive them of timely outcomes. Thus, this conventional approach is not scalable with respect to both the volume and variety of geoscience data. A much more viable solution is to couple analysis and storage systems to minimize data transfer. In our study, we compare loosely coupled approaches (exemplified by Spark and Hadoop) and tightly coupled approaches (exemplified by parallel distributed database management systems, e.g., SciDB). In particular, we investigate the optimization of data placement and movement to effectively tackle the variety challenge, and boost the popularization of parallelization to address the volume challenge. Our goal is to enable high-performance interactive analysis for a good portion of geoscience data analysis exercise. We show that tightly coupled approaches can concentrate data traffic between local storage systems and compute units, and thereby optimizing bandwidth utilization to achieve a better throughput. Based on our observations, we develop a geoscience data analysis system that tightly couples analysis engines with storages, which has direct access to the detailed map of data partition locations. Through an innovation data partitioning and distribution scheme, our system has demonstrated scalable and interactive performance in real-world geoscience data analysis applications.

Optimized distributed systems achieve significant performance improvement on sorted merging of massive VCF files.

PubMed

Sun, Xiaobo; Gao, Jingjing; Jin, Peng; Eng, Celeste; Burchard, Esteban G; Beaty, Terri H; Ruczinski, Ingo; Mathias, Rasika A; Barnes, Kathleen; Wang, Fusheng; Qin, Zhaohui S

2018-06-01

Sorted merging of genomic data is a common data operation necessary in many sequencing-based studies. It involves sorting and merging genomic data from different subjects by their genomic locations. In particular, merging a large number of variant call format (VCF) files is frequently required in large-scale whole-genome sequencing or whole-exome sequencing projects. Traditional single-machine based methods become increasingly inefficient when processing large numbers of files due to the excessive computation time and Input/Output bottleneck. Distributed systems and more recent cloud-based systems offer an attractive solution. However, carefully designed and optimized workflow patterns and execution plans (schemas) are required to take full advantage of the increased computing power while overcoming bottlenecks to achieve high performance. In this study, we custom-design optimized schemas for three Apache big data platforms, Hadoop (MapReduce), HBase, and Spark, to perform sorted merging of a large number of VCF files. These schemas all adopt the divide-and-conquer strategy to split the merging job into sequential phases/stages consisting of subtasks that are conquered in an ordered, parallel, and bottleneck-free way. In two illustrating examples, we test the performance of our schemas on merging multiple VCF files into either a single TPED or a single VCF file, which are benchmarked with the traditional single/parallel multiway-merge methods, message passing interface (MPI)-based high-performance computing (HPC) implementation, and the popular VCFTools. Our experiments suggest all three schemas either deliver a significant improvement in efficiency or render much better strong and weak scalabilities over traditional methods. Our findings provide generalized scalable schemas for performing sorted merging on genetics and genomics data using these Apache distributed systems.

Optimized distributed systems achieve significant performance improvement on sorted merging of massive VCF files

PubMed Central

Gao, Jingjing; Jin, Peng; Eng, Celeste; Burchard, Esteban G; Beaty, Terri H; Ruczinski, Ingo; Mathias, Rasika A; Barnes, Kathleen; Wang, Fusheng

2018-01-01

Abstract Background Sorted merging of genomic data is a common data operation necessary in many sequencing-based studies. It involves sorting and merging genomic data from different subjects by their genomic locations. In particular, merging a large number of variant call format (VCF) files is frequently required in large-scale whole-genome sequencing or whole-exome sequencing projects. Traditional single-machine based methods become increasingly inefficient when processing large numbers of files due to the excessive computation time and Input/Output bottleneck. Distributed systems and more recent cloud-based systems offer an attractive solution. However, carefully designed and optimized workflow patterns and execution plans (schemas) are required to take full advantage of the increased computing power while overcoming bottlenecks to achieve high performance. Findings In this study, we custom-design optimized schemas for three Apache big data platforms, Hadoop (MapReduce), HBase, and Spark, to perform sorted merging of a large number of VCF files. These schemas all adopt the divide-and-conquer strategy to split the merging job into sequential phases/stages consisting of subtasks that are conquered in an ordered, parallel, and bottleneck-free way. In two illustrating examples, we test the performance of our schemas on merging multiple VCF files into either a single TPED or a single VCF file, which are benchmarked with the traditional single/parallel multiway-merge methods, message passing interface (MPI)–based high-performance computing (HPC) implementation, and the popular VCFTools. Conclusions Our experiments suggest all three schemas either deliver a significant improvement in efficiency or render much better strong and weak scalabilities over traditional methods. Our findings provide generalized scalable schemas for performing sorted merging on genetics and genomics data using these Apache distributed systems. PMID:29762754

Advanced Computational Methods for Security Constrained Financial Transmission Rights: Structure and Parallelism

DOE Office of Scientific and Technical Information (OSTI.GOV)

Elbert, Stephen T.; Kalsi, Karanjit; Vlachopoulou, Maria

Financial Transmission Rights (FTRs) help power market participants reduce price risks associated with transmission congestion. FTRs are issued based on a process of solving a constrained optimization problem with the objective to maximize the FTR social welfare under power flow security constraints. Security constraints for different FTR categories (monthly, seasonal or annual) are usually coupled and the number of constraints increases exponentially with the number of categories. Commercial software for FTR calculation can only provide limited categories of FTRs due to the inherent computational challenges mentioned above. In this paper, a novel non-linear dynamical system (NDS) approach is proposed tomore » solve the optimization problem. The new formulation and performance of the NDS solver is benchmarked against widely used linear programming (LP) solvers like CPLEX™ and tested on large-scale systems using data from the Western Electricity Coordinating Council (WECC). The NDS is demonstrated to outperform the widely used CPLEX algorithms while exhibiting superior scalability. Furthermore, the NDS based solver can be easily parallelized which results in significant computational improvement.« less

Scalable and fast heterogeneous molecular simulation with predictive parallelization schemes

NASA Astrophysics Data System (ADS)

Guzman, Horacio V.; Junghans, Christoph; Kremer, Kurt; Stuehn, Torsten

2017-11-01

Multiscale and inhomogeneous molecular systems are challenging topics in the field of molecular simulation. In particular, modeling biological systems in the context of multiscale simulations and exploring material properties are driving a permanent development of new simulation methods and optimization algorithms. In computational terms, those methods require parallelization schemes that make a productive use of computational resources for each simulation and from its genesis. Here, we introduce the heterogeneous domain decomposition approach, which is a combination of an heterogeneity-sensitive spatial domain decomposition with an a priori rearrangement of subdomain walls. Within this approach, the theoretical modeling and scaling laws for the force computation time are proposed and studied as a function of the number of particles and the spatial resolution ratio. We also show the new approach capabilities, by comparing it to both static domain decomposition algorithms and dynamic load-balancing schemes. Specifically, two representative molecular systems have been simulated and compared to the heterogeneous domain decomposition proposed in this work. These two systems comprise an adaptive resolution simulation of a biomolecule solvated in water and a phase-separated binary Lennard-Jones fluid.

Simplified Parallel Domain Traversal

DOE Office of Scientific and Technical Information (OSTI.GOV)

Erickson III, David J

2011-01-01

Many data-intensive scientific analysis techniques require global domain traversal, which over the years has been a bottleneck for efficient parallelization across distributed-memory architectures. Inspired by MapReduce and other simplified parallel programming approaches, we have designed DStep, a flexible system that greatly simplifies efficient parallelization of domain traversal techniques at scale. In order to deliver both simplicity to users as well as scalability on HPC platforms, we introduce a novel two-tiered communication architecture for managing and exploiting asynchronous communication loads. We also integrate our design with advanced parallel I/O techniques that operate directly on native simulation output. We demonstrate DStep bymore » performing teleconnection analysis across ensemble runs of terascale atmospheric CO{sub 2} and climate data, and we show scalability results on up to 65,536 IBM BlueGene/P cores.« less

Parallelizing alternating direction implicit solver on GPUs

USDA-ARS?s Scientific Manuscript database

We present a parallel Alternating Direction Implicit (ADI) solver on GPUs. Our implementation significantly improves existing implementations in two aspects. First, we address the scalability issue of existing Parallel Cyclic Reduction (PCR) implementations by eliminating their hardware resource con...

SediFoam: A general-purpose, open-source CFD-DEM solver for particle-laden flow with emphasis on sediment transport

NASA Astrophysics Data System (ADS)

Sun, Rui; Xiao, Heng

2016-04-01

With the growth of available computational resource, CFD-DEM (computational fluid dynamics-discrete element method) becomes an increasingly promising and feasible approach for the study of sediment transport. Several existing CFD-DEM solvers are applied in chemical engineering and mining industry. However, a robust CFD-DEM solver for the simulation of sediment transport is still desirable. In this work, the development of a three-dimensional, massively parallel, and open-source CFD-DEM solver SediFoam is detailed. This solver is built based on open-source solvers OpenFOAM and LAMMPS. OpenFOAM is a CFD toolbox that can perform three-dimensional fluid flow simulations on unstructured meshes; LAMMPS is a massively parallel DEM solver for molecular dynamics. Several validation tests of SediFoam are performed using cases of a wide range of complexities. The results obtained in the present simulations are consistent with those in the literature, which demonstrates the capability of SediFoam for sediment transport applications. In addition to the validation test, the parallel efficiency of SediFoam is studied to test the performance of the code for large-scale and complex simulations. The parallel efficiency tests show that the scalability of SediFoam is satisfactory in the simulations using up to O(107) particles.

Network selection, Information filtering and Scalable computation

NASA Astrophysics Data System (ADS)

Ye, Changqing

This dissertation explores two application scenarios of sparsity pursuit method on large scale data sets. The first scenario is classification and regression in analyzing high dimensional structured data, where predictors corresponds to nodes of a given directed graph. This arises in, for instance, identification of disease genes for the Parkinson's diseases from a network of candidate genes. In such a situation, directed graph describes dependencies among the genes, where direction of edges represent certain causal effects. Key to high-dimensional structured classification and regression is how to utilize dependencies among predictors as specified by directions of the graph. In this dissertation, we develop a novel method that fully takes into account such dependencies formulated through certain nonlinear constraints. We apply the proposed method to two applications, feature selection in large margin binary classification and in linear regression. We implement the proposed method through difference convex programming for the cost function and constraints. Finally, theoretical and numerical analyses suggest that the proposed method achieves the desired objectives. An application to disease gene identification is presented. The second application scenario is personalized information filtering which extracts the information specifically relevant to a user, predicting his/her preference over a large number of items, based on the opinions of users who think alike or its content. This problem is cast into the framework of regression and classification, where we introduce novel partial latent models to integrate additional user-specific and content-specific predictors, for higher predictive accuracy. In particular, we factorize a user-over-item preference matrix into a product of two matrices, each representing a user's preference and an item preference by users. Then we propose a likelihood method to seek a sparsest latent factorization, from a class of over-complete factorizations, possibly with a high percentage of missing values. This promotes additional sparsity beyond rank reduction. Computationally, we design methods based on a ``decomposition and combination'' strategy, to break large-scale optimization into many small subproblems to solve in a recursive and parallel manner. On this basis, we implement the proposed methods through multi-platform shared-memory parallel programming, and through Mahout, a library for scalable machine learning and data mining, for mapReduce computation. For example, our methods are scalable to a dataset consisting of three billions of observations on a single machine with sufficient memory, having good timings. Both theoretical and numerical investigations show that the proposed methods exhibit significant improvement in accuracy over state-of-the-art scalable methods.

Bioinformatics algorithm based on a parallel implementation of a machine learning approach using transducers

NASA Astrophysics Data System (ADS)

Roche-Lima, Abiel; Thulasiram, Ruppa K.

2012-02-01

Finite automata, in which each transition is augmented with an output label in addition to the familiar input label, are considered finite-state transducers. Transducers have been used to analyze some fundamental issues in bioinformatics. Weighted finite-state transducers have been proposed to pairwise alignments of DNA and protein sequences; as well as to develop kernels for computational biology. Machine learning algorithms for conditional transducers have been implemented and used for DNA sequence analysis. Transducer learning algorithms are based on conditional probability computation. It is calculated by using techniques, such as pair-database creation, normalization (with Maximum-Likelihood normalization) and parameters optimization (with Expectation-Maximization - EM). These techniques are intrinsically costly for computation, even worse when are applied to bioinformatics, because the databases sizes are large. In this work, we describe a parallel implementation of an algorithm to learn conditional transducers using these techniques. The algorithm is oriented to bioinformatics applications, such as alignments, phylogenetic trees, and other genome evolution studies. Indeed, several experiences were developed using the parallel and sequential algorithm on Westgrid (specifically, on the Breeze cluster). As results, we obtain that our parallel algorithm is scalable, because execution times are reduced considerably when the data size parameter is increased. Another experience is developed by changing precision parameter. In this case, we obtain smaller execution times using the parallel algorithm. Finally, number of threads used to execute the parallel algorithm on the Breezy cluster is changed. In this last experience, we obtain as result that speedup is considerably increased when more threads are used; however there is a convergence for number of threads equal to or greater than 16.

A highly efficient 3D level-set grain growth algorithm tailored for ccNUMA architecture

NASA Astrophysics Data System (ADS)

Mießen, C.; Velinov, N.; Gottstein, G.; Barrales-Mora, L. A.

2017-12-01

A highly efficient simulation model for 2D and 3D grain growth was developed based on the level-set method. The model introduces modern computational concepts to achieve excellent performance on parallel computer architectures. Strong scalability was measured on cache-coherent non-uniform memory access (ccNUMA) architectures. To achieve this, the proposed approach considers the application of local level-set functions at the grain level. Ideal and non-ideal grain growth was simulated in 3D with the objective to study the evolution of statistical representative volume elements in polycrystals. In addition, microstructure evolution in an anisotropic magnetic material affected by an external magnetic field was simulated.

Achieving production-level use of HEP software at the Argonne Leadership Computing Facility

NASA Astrophysics Data System (ADS)

Uram, T. D.; Childers, J. T.; LeCompte, T. J.; Papka, M. E.; Benjamin, D.

2015-12-01

HEP's demand for computing resources has grown beyond the capacity of the Grid, and these demands will accelerate with the higher energy and luminosity planned for Run II. Mira, the ten petaFLOPs supercomputer at the Argonne Leadership Computing Facility, is a potentially significant compute resource for HEP research. Through an award of fifty million hours on Mira, we have delivered millions of events to LHC experiments by establishing the means of marshaling jobs through serial stages on local clusters, and parallel stages on Mira. We are running several HEP applications, including Alpgen, Pythia, Sherpa, and Geant4. Event generators, such as Sherpa, typically have a split workload: a small scale integration phase, and a second, more scalable, event-generation phase. To accommodate this workload on Mira we have developed two Python-based Django applications, Balsam and ARGO. Balsam is a generalized scheduler interface which uses a plugin system for interacting with scheduler software such as HTCondor, Cobalt, and TORQUE. ARGO is a workflow manager that submits jobs to instances of Balsam. Through these mechanisms, the serial and parallel tasks within jobs are executed on the appropriate resources. This approach and its integration with the PanDA production system will be discussed.

Toward Petascale Biologically Plausible Neural Networks

NASA Astrophysics Data System (ADS)

Long, Lyle

This talk will describe an approach to achieving petascale neural networks. Artificial intelligence has been oversold for many decades. Computers in the beginning could only do about 16,000 operations per second. Computer processing power, however, has been doubling every two years thanks to Moore's law, and growing even faster due to massively parallel architectures. Finally, 60 years after the first AI conference we have computers on the order of the performance of the human brain (1016 operations per second). The main issues now are algorithms, software, and learning. We have excellent models of neurons, such as the Hodgkin-Huxley model, but we do not know how the human neurons are wired together. With careful attention to efficient parallel computing, event-driven programming, table lookups, and memory minimization massive scale simulations can be performed. The code that will be described was written in C + + and uses the Message Passing Interface (MPI). It uses the full Hodgkin-Huxley neuron model, not a simplified model. It also allows arbitrary network structures (deep, recurrent, convolutional, all-to-all, etc.). The code is scalable, and has, so far, been tested on up to 2,048 processor cores using 107 neurons and 109 synapses.

PFLOTRAN User Manual: A Massively Parallel Reactive Flow and Transport Model for Describing Surface and Subsurface Processes

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lichtner, Peter C.; Hammond, Glenn E.; Lu, Chuan

PFLOTRAN solves a system of generally nonlinear partial differential equations describing multi-phase, multicomponent and multiscale reactive flow and transport in porous materials. The code is designed to run on massively parallel computing architectures as well as workstations and laptops (e.g. Hammond et al., 2011). Parallelization is achieved through domain decomposition using the PETSc (Portable Extensible Toolkit for Scientific Computation) libraries for the parallelization framework (Balay et al., 1997). PFLOTRAN has been developed from the ground up for parallel scalability and has been run on up to 218 processor cores with problem sizes up to 2 billion degrees of freedom. Writtenmore » in object oriented Fortran 90, the code requires the latest compilers compatible with Fortran 2003. At the time of this writing this requires gcc 4.7.x, Intel 12.1.x and PGC compilers. As a requirement of running problems with a large number of degrees of freedom, PFLOTRAN allows reading input data that is too large to fit into memory allotted to a single processor core. The current limitation to the problem size PFLOTRAN can handle is the limitation of the HDF5 file format used for parallel IO to 32 bit integers. Noting that 2 32 = 4; 294; 967; 296, this gives an estimate of the maximum problem size that can be currently run with PFLOTRAN. Hopefully this limitation will be remedied in the near future.« less

GPU acceleration of a petascale application for turbulent mixing at high Schmidt number using OpenMP 4.5

NASA Astrophysics Data System (ADS)

Clay, M. P.; Buaria, D.; Yeung, P. K.; Gotoh, T.

2018-07-01

This paper reports on the successful implementation of a massively parallel GPU-accelerated algorithm for the direct numerical simulation of turbulent mixing at high Schmidt number. The work stems from a recent development (Comput. Phys. Commun., vol. 219, 2017, 313-328), in which a low-communication algorithm was shown to attain high degrees of scalability on the Cray XE6 architecture when overlapping communication and computation via dedicated communication threads. An even higher level of performance has now been achieved using OpenMP 4.5 on the Cray XK7 architecture, where on each node the 16 integer cores of an AMD Interlagos processor share a single Nvidia K20X GPU accelerator. In the new algorithm, data movements are minimized by performing virtually all of the intensive scalar field computations in the form of combined compact finite difference (CCD) operations on the GPUs. A memory layout in departure from usual practices is found to provide much better performance for a specific kernel required to apply the CCD scheme. Asynchronous execution enabled by adding the OpenMP 4.5 NOWAIT clause to TARGET constructs improves scalability when used to overlap computation on the GPUs with computation and communication on the CPUs. On the 27-petaflops supercomputer Titan at Oak Ridge National Laboratory, USA, a GPU-to-CPU speedup factor of approximately 5 is consistently observed at the largest problem size of 81923 grid points for the scalar field computed with 8192 XK7 nodes.

Scalable parallel elastic-plastic finite element analysis using a quasi-Newton method with a balancing domain decomposition preconditioner

NASA Astrophysics Data System (ADS)

Yusa, Yasunori; Okada, Hiroshi; Yamada, Tomonori; Yoshimura, Shinobu

2018-04-01

A domain decomposition method for large-scale elastic-plastic problems is proposed. The proposed method is based on a quasi-Newton method in conjunction with a balancing domain decomposition preconditioner. The use of a quasi-Newton method overcomes two problems associated with the conventional domain decomposition method based on the Newton-Raphson method: (1) avoidance of a double-loop iteration algorithm, which generally has large computational complexity, and (2) consideration of the local concentration of nonlinear deformation, which is observed in elastic-plastic problems with stress concentration. Moreover, the application of a balancing domain decomposition preconditioner ensures scalability. Using the conventional and proposed domain decomposition methods, several numerical tests, including weak scaling tests, were performed. The convergence performance of the proposed method is comparable to that of the conventional method. In particular, in elastic-plastic analysis, the proposed method exhibits better convergence performance than the conventional method.

A Multibody Formulation for Three Dimensional Brick Finite Element Based Parallel and Scalable Rotor Dynamic Analysis

DTIC Science & Technology

2010-05-01

connections near the hub end, and containing up to 0.48 million degrees of freedom. The models are analyzed for scala - bility and timing for hover and...Parallel and Scalable Rotor Dynamic Analysis 5a. CONTRACT NUMBER 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER 5e. TASK...will enable the modeling of critical couplings that occur in hingeless and bearingless hubs with advanced flex structures. Second , it will enable the

Polyphony: A Workflow Orchestration Framework for Cloud Computing

NASA Technical Reports Server (NTRS)

Shams, Khawaja S.; Powell, Mark W.; Crockett, Tom M.; Norris, Jeffrey S.; Rossi, Ryan; Soderstrom, Tom

2010-01-01

Cloud Computing has delivered unprecedented compute capacity to NASA missions at affordable rates. Missions like the Mars Exploration Rovers (MER) and Mars Science Lab (MSL) are enjoying the elasticity that enables them to leverage hundreds, if not thousands, or machines for short durations without making any hardware procurements. In this paper, we describe Polyphony, a resilient, scalable, and modular framework that efficiently leverages a large set of computing resources to perform parallel computations. Polyphony can employ resources on the cloud, excess capacity on local machines, as well as spare resources on the supercomputing center, and it enables these resources to work in concert to accomplish a common goal. Polyphony is resilient to node failures, even if they occur in the middle of a transaction. We will conclude with an evaluation of a production-ready application built on top of Polyphony to perform image-processing operations of images from around the solar system, including Mars, Saturn, and Titan.

Adagio 4.20 User’s Guide

DOE Office of Scientific and Technical Information (OSTI.GOV)

Spencer, Benjamin Whiting; Crane, Nathan K.; Heinstein, Martin W.

2011-03-01

Adagio is a Lagrangian, three-dimensional, implicit code for the analysis of solids and structures. It uses a multi-level iterative solver, which enables it to solve problems with large deformations, nonlinear material behavior, and contact. It also has a versatile library of continuum and structural elements, and an extensive library of material models. Adagio is written for parallel computing environments, and its solvers allow for scalable solutions of very large problems. Adagio uses the SIERRA Framework, which allows for coupling with other SIERRA mechanics codes. This document describes the functionality and input structure for Adagio.

a Hadoop-Based Distributed Framework for Efficient Managing and Processing Big Remote Sensing Images

NASA Astrophysics Data System (ADS)

Wang, C.; Hu, F.; Hu, X.; Zhao, S.; Wen, W.; Yang, C.

2015-07-01

Various sensors from airborne and satellite platforms are producing large volumes of remote sensing images for mapping, environmental monitoring, disaster management, military intelligence, and others. However, it is challenging to efficiently storage, query and process such big data due to the data- and computing- intensive issues. In this paper, a Hadoop-based framework is proposed to manage and process the big remote sensing data in a distributed and parallel manner. Especially, remote sensing data can be directly fetched from other data platforms into the Hadoop Distributed File System (HDFS). The Orfeo toolbox, a ready-to-use tool for large image processing, is integrated into MapReduce to provide affluent image processing operations. With the integration of HDFS, Orfeo toolbox and MapReduce, these remote sensing images can be directly processed in parallel in a scalable computing environment. The experiment results show that the proposed framework can efficiently manage and process such big remote sensing data.

Development and Application of a Parallel LCAO Cluster Method

NASA Astrophysics Data System (ADS)

Patton, David C.

1997-08-01

CPU intensive steps in the SCF electronic structure calculations of clusters and molecules with a first-principles LCAO method have been fully parallelized via a message passing paradigm. Identification of the parts of the code that are composed of many independent compute-intensive steps is discussed in detail as they are the most readily parallelized. Most of the parallelization involves spatially decomposing numerical operations on a mesh. One exception is the solution of Poisson's equation which relies on distribution of the charge density and multipole methods. The method we use to parallelize this part of the calculation is quite novel and is covered in detail. We present a general method for dynamically load-balancing a parallel calculation and discuss how we use this method in our code. The results of benchmark calculations of the IR and Raman spectra of PAH molecules such as anthracene (C_14H_10) and tetracene (C_18H_12) are presented. These benchmark calculations were performed on an IBM SP2 and a SUN Ultra HPC server with both MPI and PVM. Scalability and speedup for these calculations is analyzed to determine the efficiency of the code. In addition, performance and usage issues for MPI and PVM are presented.

WESTPA: An interoperable, highly scalable software package for weighted ensemble simulation and analysis

PubMed Central

Zwier, Matthew C.; Adelman, Joshua L.; Kaus, Joseph W.; Pratt, Adam J.; Wong, Kim F.; Rego, Nicholas B.; Suárez, Ernesto; Lettieri, Steven; Wang, David W.; Grabe, Michael; Zuckerman, Daniel M.; Chong, Lillian T.

2015-01-01

The weighted ensemble (WE) path sampling approach orchestrates an ensemble of parallel calculations with intermittent communication to enhance the sampling of rare events, such as molecular associations or conformational changes in proteins or peptides. Trajectories are replicated and pruned in a way that focuses computational effort on under-explored regions of configuration space while maintaining rigorous kinetics. To enable the simulation of rare events at any scale (e.g. atomistic, cellular), we have developed an open-source, interoperable, and highly scalable software package for the execution and analysis of WE simulations: WESTPA (The Weighted Ensemble Simulation Toolkit with Parallelization and Analysis). WESTPA scales to thousands of CPU cores and includes a suite of analysis tools that have been implemented in a massively parallel fashion. The software has been designed to interface conveniently with any dynamics engine and has already been used with a variety of molecular dynamics (e.g. GROMACS, NAMD, OpenMM, AMBER) and cell-modeling packages (e.g. BioNetGen, MCell). WESTPA has been in production use for over a year, and its utility has been demonstrated for a broad set of problems, ranging from atomically detailed host-guest associations to non-spatial chemical kinetics of cellular signaling networks. The following describes the design and features of WESTPA, including the facilities it provides for running WE simulations, storing and analyzing WE simulation data, as well as examples of input and output. PMID:26392815

Hybrid-optimization strategy for the communication of large-scale Kinetic Monte Carlo simulation

NASA Astrophysics Data System (ADS)

Wu, Baodong; Li, Shigang; Zhang, Yunquan; Nie, Ningming

2017-02-01

The parallel Kinetic Monte Carlo (KMC) algorithm based on domain decomposition has been widely used in large-scale physical simulations. However, the communication overhead of the parallel KMC algorithm is critical, and severely degrades the overall performance and scalability. In this paper, we present a hybrid optimization strategy to reduce the communication overhead for the parallel KMC simulations. We first propose a communication aggregation algorithm to reduce the total number of messages and eliminate the communication redundancy. Then, we utilize the shared memory to reduce the memory copy overhead of the intra-node communication. Finally, we optimize the communication scheduling using the neighborhood collective operations. We demonstrate the scalability and high performance of our hybrid optimization strategy by both theoretical and experimental analysis. Results show that the optimized KMC algorithm exhibits better performance and scalability than the well-known open-source library-SPPARKS. On 32-node Xeon E5-2680 cluster (total 640 cores), the optimized algorithm reduces the communication time by 24.8% compared with SPPARKS.

Numerically stable, scalable formulas for parallel and online computation of higher-order multivariate central moments with arbitrary weights

DOE PAGES

Pebay, Philippe; Terriberry, Timothy B.; Kolla, Hemanth; ...

2016-03-29

Formulas for incremental or parallel computation of second order central moments have long been known, and recent extensions of these formulas to univariate and multivariate moments of arbitrary order have been developed. Such formulas are of key importance in scenarios where incremental results are required and in parallel and distributed systems where communication costs are high. We survey these recent results, and improve them with arbitrary-order, numerically stable one-pass formulas which we further extend with weighted and compound variants. We also develop a generalized correction factor for standard two-pass algorithms that enables the maintenance of accuracy over nearly the fullmore » representable range of the input, avoiding the need for extended-precision arithmetic. We then empirically examine algorithm correctness for pairwise update formulas up to order four as well as condition number and relative error bounds for eight different central moment formulas, each up to degree six, to address the trade-offs between numerical accuracy and speed of the various algorithms. Finally, we demonstrate the use of the most elaborate among the above mentioned formulas, with the utilization of the compound moments for a practical large-scale scientific application.« less

Accelerating k-NN Algorithm with Hybrid MPI and OpenSHMEM

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lin, Jian; Hamidouche, Khaled; Zheng, Jie

2015-08-05

Machine Learning algorithms are benefiting from the continuous improvement of programming models, including MPI, MapReduce and PGAS. k-Nearest Neighbors (k-NN) algorithm is a widely used machine learning algorithm, applied to supervised learning tasks such as classification. Several parallel implementations of k-NN have been proposed in the literature and practice. However, on high-performance computing systems with high-speed interconnects, it is important to further accelerate existing designs of the k-NN algorithm through taking advantage of scalable programming models. To improve the performance of k-NN on large-scale environment with InfiniBand network, this paper proposes several alternative hybrid MPI+OpenSHMEM designs and performs a systemicmore » evaluation and analysis on typical workloads. The hybrid designs leverage the one-sided memory access to better overlap communication with computation than the existing pure MPI design, and propose better schemes for efficient buffer management. The implementation based on k-NN program from MaTEx with MVAPICH2-X (Unified MPI+PGAS Communication Runtime over InfiniBand) shows up to 9.0% time reduction for training KDD Cup 2010 workload over 512 cores, and 27.6% time reduction for small workload with balanced communication and computation. Experiments of running with varied number of cores show that our design can maintain good scalability.« less

Fast MPEG-CDVS Encoder With GPU-CPU Hybrid Computing.

PubMed

Duan, Ling-Yu; Sun, Wei; Zhang, Xinfeng; Wang, Shiqi; Chen, Jie; Yin, Jianxiong; See, Simon; Huang, Tiejun; Kot, Alex C; Gao, Wen

2018-05-01

The compact descriptors for visual search (CDVS) standard from ISO/IEC moving pictures experts group has succeeded in enabling the interoperability for efficient and effective image retrieval by standardizing the bitstream syntax of compact feature descriptors. However, the intensive computation of a CDVS encoder unfortunately hinders its widely deployment in industry for large-scale visual search. In this paper, we revisit the merits of low complexity design of CDVS core techniques and present a very fast CDVS encoder by leveraging the massive parallel execution resources of graphics processing unit (GPU). We elegantly shift the computation-intensive and parallel-friendly modules to the state-of-the-arts GPU platforms, in which the thread block allocation as well as the memory access mechanism are jointly optimized to eliminate performance loss. In addition, those operations with heavy data dependence are allocated to CPU for resolving the extra but non-necessary computation burden for GPU. Furthermore, we have demonstrated the proposed fast CDVS encoder can work well with those convolution neural network approaches which enables to leverage the advantages of GPU platforms harmoniously, and yield significant performance improvements. Comprehensive experimental results over benchmarks are evaluated, which has shown that the fast CDVS encoder using GPU-CPU hybrid computing is promising for scalable visual search.

DInSAR time series generation within a cloud computing environment: from ERS to Sentinel-1 scenario

NASA Astrophysics Data System (ADS)

Casu, Francesco; Elefante, Stefano; Imperatore, Pasquale; Lanari, Riccardo; Manunta, Michele; Zinno, Ivana; Mathot, Emmanuel; Brito, Fabrice; Farres, Jordi; Lengert, Wolfgang

2013-04-01

One of the techniques that will strongly benefit from the advent of the Sentinel-1 system is Differential SAR Interferometry (DInSAR), which has successfully demonstrated to be an effective tool to detect and monitor ground displacements with centimetre accuracy. The geoscience communities (volcanology, seismicity, …), as well as those related to hazard monitoring and risk mitigation, make extensively use of the DInSAR technique and they will take advantage from the huge amount of SAR data acquired by Sentinel-1. Indeed, such an information will successfully permit the generation of Earth's surface displacement maps and time series both over large areas and long time span. However, the issue of managing, processing and analysing the large Sentinel data stream is envisaged by the scientific community to be a major bottleneck, particularly during crisis phases. The emerging need of creating a common ecosystem in which data, results and processing tools are shared, is envisaged to be a successful way to address such a problem and to contribute to the information and knowledge spreading. The Supersites initiative as well as the ESA SuperSites Exploitation Platform (SSEP) and the ESA Cloud Computing Operational Pilot (CIOP) projects provide effective answers to this need and they are pushing towards the development of such an ecosystem. It is clear that all the current and existent tools for querying, processing and analysing SAR data are required to be not only updated for managing the large data stream of Sentinel-1 satellite, but also reorganized for quickly replying to the simultaneous and highly demanding user requests, mainly during emergency situations. This translates into the automatic and unsupervised processing of large amount of data as well as the availability of scalable, widely accessible and high performance computing capabilities. The cloud computing environment permits to achieve all of these objectives, particularly in case of spike and peak requests of processing resources linked to disaster events. This work aims at presenting a parallel computational model for the widely used DInSAR algorithm named as Small BAseline Subset (SBAS), which has been implemented within the cloud computing environment provided by the ESA-CIOP platform. This activity has resulted in developing a scalable, unsupervised, portable, and widely accessible (through a web portal) parallel DInSAR computational tool. The activity has rewritten and developed the SBAS application algorithm within a parallel system environment, i.e., in a form that allows us to benefit from multiple processing units. This requires the devising a parallel version of the SBAS algorithm and its subsequent implementation, implying additional complexity in algorithm designing and an efficient multi processor programming, with the final aim of a parallel performance optimization. Although the presented algorithm has been designed to work with Sentinel-1 data, it can also process other satellite SAR data (ERS, ENVISAT, CSK, TSX, ALOS). Indeed, the performance analysis of the implemented SBAS parallel version has been tested on the full ASAR archive (64 acquisitions) acquired over the Napoli Bay, a volcanic and densely urbanized area in Southern Italy. The full processing - from the raw data download to the generation of DInSAR time series - has been carried out by engaging 4 nodes, each one with 2 cores and 16 GB of RAM, and has taken about 36 hours, with respect to about 135 hours of the sequential version. Extensive analysis on other test areas significant from DInSAR and geophysical viewpoint will be presented. Finally, preliminary performance evaluation of the presented approach within the Sentinel-1 scenario will be provided.

Real-time computing platform for spiking neurons (RT-spike).

PubMed

Ros, Eduardo; Ortigosa, Eva M; Agís, Rodrigo; Carrillo, Richard; Arnold, Michael

2006-07-01

A computing platform is described for simulating arbitrary networks of spiking neurons in real time. A hybrid computing scheme is adopted that uses both software and hardware components to manage the tradeoff between flexibility and computational power; the neuron model is implemented in hardware and the network model and the learning are implemented in software. The incremental transition of the software components into hardware is supported. We focus on a spike response model (SRM) for a neuron where the synapses are modeled as input-driven conductances. The temporal dynamics of the synaptic integration process are modeled with a synaptic time constant that results in a gradual injection of charge. This type of model is computationally expensive and is not easily amenable to existing software-based event-driven approaches. As an alternative we have designed an efficient time-based computing architecture in hardware, where the different stages of the neuron model are processed in parallel. Further improvements occur by computing multiple neurons in parallel using multiple processing units. This design is tested using reconfigurable hardware and its scalability and performance evaluated. Our overall goal is to investigate biologically realistic models for the real-time control of robots operating within closed action-perception loops, and so we evaluate the performance of the system on simulating a model of the cerebellum where the emulation of the temporal dynamics of the synaptic integration process is important.

Scalable direct Vlasov solver with discontinuous Galerkin method on unstructured mesh.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Xu, J.; Ostroumov, P. N.; Mustapha, B.

2010-12-01

This paper presents the development of parallel direct Vlasov solvers with discontinuous Galerkin (DG) method for beam and plasma simulations in four dimensions. Both physical and velocity spaces are in two dimesions (2P2V) with unstructured mesh. Contrary to the standard particle-in-cell (PIC) approach for kinetic space plasma simulations, i.e., solving Vlasov-Maxwell equations, direct method has been used in this paper. There are several benefits to solving a Vlasov equation directly, such as avoiding noise associated with a finite number of particles and the capability to capture fine structure in the plasma. The most challanging part of a direct Vlasov solvermore » comes from higher dimensions, as the computational cost increases as N{sup 2d}, where d is the dimension of the physical space. Recently, due to the fast development of supercomputers, the possibility has become more realistic. Many efforts have been made to solve Vlasov equations in low dimensions before; now more interest has focused on higher dimensions. Different numerical methods have been tried so far, such as the finite difference method, Fourier Spectral method, finite volume method, and spectral element method. This paper is based on our previous efforts to use the DG method. The DG method has been proven to be very successful in solving Maxwell equations, and this paper is our first effort in applying the DG method to Vlasov equations. DG has shown several advantages, such as local mass matrix, strong stability, and easy parallelization. These are particularly suitable for Vlasov equations. Domain decomposition in high dimensions has been used for parallelization; these include a highly scalable parallel two-dimensional Poisson solver. Benchmark results have been shown and simulation results will be reported.« less

Comparing Neuromorphic Solutions in Action: Implementing a Bio-Inspired Solution to a Benchmark Classification Task on Three Parallel-Computing Platforms

PubMed Central

Diamond, Alan; Nowotny, Thomas; Schmuker, Michael

2016-01-01

Neuromorphic computing employs models of neuronal circuits to solve computing problems. Neuromorphic hardware systems are now becoming more widely available and “neuromorphic algorithms” are being developed. As they are maturing toward deployment in general research environments, it becomes important to assess and compare them in the context of the applications they are meant to solve. This should encompass not just task performance, but also ease of implementation, speed of processing, scalability, and power efficiency. Here, we report our practical experience of implementing a bio-inspired, spiking network for multivariate classification on three different platforms: the hybrid digital/analog Spikey system, the digital spike-based SpiNNaker system, and GeNN, a meta-compiler for parallel GPU hardware. We assess performance using a standard hand-written digit classification task. We found that whilst a different implementation approach was required for each platform, classification performances remained in line. This suggests that all three implementations were able to exercise the model's ability to solve the task rather than exposing inherent platform limits, although differences emerged when capacity was approached. With respect to execution speed and power consumption, we found that for each platform a large fraction of the computing time was spent outside of the neuromorphic device, on the host machine. Time was spent in a range of combinations of preparing the model, encoding suitable input spiking data, shifting data, and decoding spike-encoded results. This is also where a large proportion of the total power was consumed, most markedly for the SpiNNaker and Spikey systems. We conclude that the simulation efficiency advantage of the assessed specialized hardware systems is easily lost in excessive host-device communication, or non-neuronal parts of the computation. These results emphasize the need to optimize the host-device communication architecture for scalability, maximum throughput, and minimum latency. Moreover, our results indicate that special attention should be paid to minimize host-device communication when designing and implementing networks for efficient neuromorphic computing. PMID:26778950

Scalable Metropolis Monte Carlo for simulation of hard shapes

NASA Astrophysics Data System (ADS)

Anderson, Joshua A.; Eric Irrgang, M.; Glotzer, Sharon C.

2016-07-01

We design and implement a scalable hard particle Monte Carlo simulation toolkit (HPMC), and release it open source as part of HOOMD-blue. HPMC runs in parallel on many CPUs and many GPUs using domain decomposition. We employ BVH trees instead of cell lists on the CPU for fast performance, especially with large particle size disparity, and optimize inner loops with SIMD vector intrinsics on the CPU. Our GPU kernel proposes many trial moves in parallel on a checkerboard and uses a block-level queue to redistribute work among threads and avoid divergence. HPMC supports a wide variety of shape classes, including spheres/disks, unions of spheres, convex polygons, convex spheropolygons, concave polygons, ellipsoids/ellipses, convex polyhedra, convex spheropolyhedra, spheres cut by planes, and concave polyhedra. NVT and NPT ensembles can be run in 2D or 3D triclinic boxes. Additional integration schemes permit Frenkel-Ladd free energy computations and implicit depletant simulations. In a benchmark system of a fluid of 4096 pentagons, HPMC performs 10 million sweeps in 10 min on 96 CPU cores on XSEDE Comet. The same simulation would take 7.6 h in serial. HPMC also scales to large system sizes, and the same benchmark with 16.8 million particles runs in 1.4 h on 2048 GPUs on OLCF Titan.

Many-integrated core (MIC) technology for accelerating Monte Carlo simulation of radiation transport: A study based on the code DPM

NASA Astrophysics Data System (ADS)

Rodriguez, M.; Brualla, L.

2018-04-01

Monte Carlo simulation of radiation transport is computationally demanding to obtain reasonably low statistical uncertainties of the estimated quantities. Therefore, it can benefit in a large extent from high-performance computing. This work is aimed at assessing the performance of the first generation of the many-integrated core architecture (MIC) Xeon Phi coprocessor with respect to that of a CPU consisting of a double 12-core Xeon processor in Monte Carlo simulation of coupled electron-photonshowers. The comparison was made twofold, first, through a suite of basic tests including parallel versions of the random number generators Mersenne Twister and a modified implementation of RANECU. These tests were addressed to establish a baseline comparison between both devices. Secondly, through the p DPM code developed in this work. p DPM is a parallel version of the Dose Planning Method (DPM) program for fast Monte Carlo simulation of radiation transport in voxelized geometries. A variety of techniques addressed to obtain a large scalability on the Xeon Phi were implemented in p DPM. Maximum scalabilities of 84 . 2 × and 107 . 5 × were obtained in the Xeon Phi for simulations of electron and photon beams, respectively. Nevertheless, in none of the tests involving radiation transport the Xeon Phi performed better than the CPU. The disadvantage of the Xeon Phi with respect to the CPU owes to the low performance of the single core of the former. A single core of the Xeon Phi was more than 10 times less efficient than a single core of the CPU for all radiation transport simulations.

Exploiting Parallel R in the Cloud with SPRINT

PubMed Central

Piotrowski, M.; McGilvary, G.A.; Sloan, T. M.; Mewissen, M.; Lloyd, A.D.; Forster, T.; Mitchell, L.; Ghazal, P.; Hill, J.

2012-01-01

Background Advances in DNA Microarray devices and next-generation massively parallel DNA sequencing platforms have led to an exponential growth in data availability but the arising opportunities require adequate computing resources. High Performance Computing (HPC) in the Cloud offers an affordable way of meeting this need. Objectives Bioconductor, a popular tool for high-throughput genomic data analysis, is distributed as add-on modules for the R statistical programming language but R has no native capabilities for exploiting multi-processor architectures. SPRINT is an R package that enables easy access to HPC for genomics researchers. This paper investigates: setting up and running SPRINT-enabled genomic analyses on Amazon’s Elastic Compute Cloud (EC2), the advantages of submitting applications to EC2 from different parts of the world and, if resource underutilization can improve application performance. Methods The SPRINT parallel implementations of correlation, permutation testing, partitioning around medoids and the multi-purpose papply have been benchmarked on data sets of various size on Amazon EC2. Jobs have been submitted from both the UK and Thailand to investigate monetary differences. Results It is possible to obtain good, scalable performance but the level of improvement is dependent upon the nature of algorithm. Resource underutilization can further improve the time to result. End-user’s location impacts on costs due to factors such as local taxation. Conclusions: Although not designed to satisfy HPC requirements, Amazon EC2 and cloud computing in general provides an interesting alternative and provides new possibilities for smaller organisations with limited funds. PMID:23223611

Exploiting parallel R in the cloud with SPRINT.

PubMed

Piotrowski, M; McGilvary, G A; Sloan, T M; Mewissen, M; Lloyd, A D; Forster, T; Mitchell, L; Ghazal, P; Hill, J

2013-01-01

Advances in DNA Microarray devices and next-generation massively parallel DNA sequencing platforms have led to an exponential growth in data availability but the arising opportunities require adequate computing resources. High Performance Computing (HPC) in the Cloud offers an affordable way of meeting this need. Bioconductor, a popular tool for high-throughput genomic data analysis, is distributed as add-on modules for the R statistical programming language but R has no native capabilities for exploiting multi-processor architectures. SPRINT is an R package that enables easy access to HPC for genomics researchers. This paper investigates: setting up and running SPRINT-enabled genomic analyses on Amazon's Elastic Compute Cloud (EC2), the advantages of submitting applications to EC2 from different parts of the world and, if resource underutilization can improve application performance. The SPRINT parallel implementations of correlation, permutation testing, partitioning around medoids and the multi-purpose papply have been benchmarked on data sets of various size on Amazon EC2. Jobs have been submitted from both the UK and Thailand to investigate monetary differences. It is possible to obtain good, scalable performance but the level of improvement is dependent upon the nature of the algorithm. Resource underutilization can further improve the time to result. End-user's location impacts on costs due to factors such as local taxation. Although not designed to satisfy HPC requirements, Amazon EC2 and cloud computing in general provides an interesting alternative and provides new possibilities for smaller organisations with limited funds.

Scalable Unix tools on parallel processors

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gropp, W.; Lusk, E.

1994-12-31

The introduction of parallel processors that run a separate copy of Unix on each process has introduced new problems in managing the user`s environment. This paper discusses some generalizations of common Unix commands for managing files (e.g. 1s) and processes (e.g. ps) that are convenient and scalable. These basic tools, just like their Unix counterparts, are text-based. We also discuss a way to use these with a graphical user interface (GUI). Some notes on the implementation are provided. Prototypes of these commands are publicly available.

A scalable parallel algorithm for multiple objective linear programs

NASA Technical Reports Server (NTRS)

Wiecek, Malgorzata M.; Zhang, Hong

1994-01-01

This paper presents an ADBASE-based parallel algorithm for solving multiple objective linear programs (MOLP's). Job balance, speedup and scalability are of primary interest in evaluating efficiency of the new algorithm. Implementation results on Intel iPSC/2 and Paragon multiprocessors show that the algorithm significantly speeds up the process of solving MOLP's, which is understood as generating all or some efficient extreme points and unbounded efficient edges. The algorithm gives specially good results for large and very large problems. Motivation and justification for solving such large MOLP's are also included.

Approximate Computing Techniques for Iterative Graph Algorithms

DOE Office of Scientific and Technical Information (OSTI.GOV)

Panyala, Ajay R.; Subasi, Omer; Halappanavar, Mahantesh

Approximate computing enables processing of large-scale graphs by trading off quality for performance. Approximate computing techniques have become critical not only due to the emergence of parallel architectures but also the availability of large scale datasets enabling data-driven discovery. Using two prototypical graph algorithms, PageRank and community detection, we present several approximate computing heuristics to scale the performance with minimal loss of accuracy. We present several heuristics including loop perforation, data caching, incomplete graph coloring and synchronization, and evaluate their efficiency. We demonstrate performance improvements of up to 83% for PageRank and up to 450x for community detection, with lowmore » impact of accuracy for both the algorithms. We expect the proposed approximate techniques will enable scalable graph analytics on data of importance to several applications in science and their subsequent adoption to scale similar graph algorithms.« less

Accelerating Climate Simulations Through Hybrid Computing

NASA Technical Reports Server (NTRS)

Zhou, Shujia; Sinno, Scott; Cruz, Carlos; Purcell, Mark

2009-01-01

Unconventional multi-core processors (e.g., IBM Cell B/E and NYIDIDA GPU) have emerged as accelerators in climate simulation. However, climate models typically run on parallel computers with conventional processors (e.g., Intel and AMD) using MPI. Connecting accelerators to this architecture efficiently and easily becomes a critical issue. When using MPI for connection, we identified two challenges: (1) identical MPI implementation is required in both systems, and; (2) existing MPI code must be modified to accommodate the accelerators. In response, we have extended and deployed IBM Dynamic Application Virtualization (DAV) in a hybrid computing prototype system (one blade with two Intel quad-core processors, two IBM QS22 Cell blades, connected with Infiniband), allowing for seamlessly offloading compute-intensive functions to remote, heterogeneous accelerators in a scalable, load-balanced manner. Currently, a climate solar radiation model running with multiple MPI processes has been offloaded to multiple Cell blades with approx.10% network overhead.

Accelerating Full Configuration Interaction Calculations for Nuclear Structure

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yang, Chao; Sternberg, Philip; Maris, Pieter

2008-04-14

One of the emerging computational approaches in nuclear physics is the full configuration interaction (FCI) method for solving the many-body nuclear Hamiltonian in a sufficiently large single-particle basis space to obtain exact answers - either directly or by extrapolation. The lowest eigenvalues and correspondingeigenvectors for very large, sparse and unstructured nuclear Hamiltonian matrices are obtained and used to evaluate additional experimental quantities. These matrices pose a significant challenge to the design and implementation of efficient and scalable algorithms for obtaining solutions on massively parallel computer systems. In this paper, we describe the computational strategies employed in a state-of-the-art FCI codemore » MFDn (Many Fermion Dynamics - nuclear) as well as techniques we recently developed to enhance the computational efficiency of MFDn. We will demonstrate the current capability of MFDn and report the latest performance improvement we have achieved. We will also outline our future research directions.« less

Fast l₁-SPIRiT compressed sensing parallel imaging MRI: scalable parallel implementation and clinically feasible runtime.

PubMed

Murphy, Mark; Alley, Marcus; Demmel, James; Keutzer, Kurt; Vasanawala, Shreyas; Lustig, Michael

2012-06-01

We present l₁-SPIRiT, a simple algorithm for auto calibrating parallel imaging (acPI) and compressed sensing (CS) that permits an efficient implementation with clinically-feasible runtimes. We propose a CS objective function that minimizes cross-channel joint sparsity in the wavelet domain. Our reconstruction minimizes this objective via iterative soft-thresholding, and integrates naturally with iterative self-consistent parallel imaging (SPIRiT). Like many iterative magnetic resonance imaging reconstructions, l₁-SPIRiT's image quality comes at a high computational cost. Excessively long runtimes are a barrier to the clinical use of any reconstruction approach, and thus we discuss our approach to efficiently parallelizing l₁-SPIRiT and to achieving clinically-feasible runtimes. We present parallelizations of l₁-SPIRiT for both multi-GPU systems and multi-core CPUs, and discuss the software optimization and parallelization decisions made in our implementation. The performance of these alternatives depends on the processor architecture, the size of the image matrix, and the number of parallel imaging channels. Fundamentally, achieving fast runtime requires the correct trade-off between cache usage and parallelization overheads. We demonstrate image quality via a case from our clinical experimentation, using a custom 3DFT spoiled gradient echo (SPGR) sequence with up to 8× acceleration via Poisson-disc undersampling in the two phase-encoded directions.

Fast ℓ1-SPIRiT Compressed Sensing Parallel Imaging MRI: Scalable Parallel Implementation and Clinically Feasible Runtime

PubMed Central

Murphy, Mark; Alley, Marcus; Demmel, James; Keutzer, Kurt; Vasanawala, Shreyas; Lustig, Michael

2012-01-01

We present ℓ1-SPIRiT, a simple algorithm for auto calibrating parallel imaging (acPI) and compressed sensing (CS) that permits an efficient implementation with clinically-feasible runtimes. We propose a CS objective function that minimizes cross-channel joint sparsity in the Wavelet domain. Our reconstruction minimizes this objective via iterative soft-thresholding, and integrates naturally with iterative Self-Consistent Parallel Imaging (SPIRiT). Like many iterative MRI reconstructions, ℓ1-SPIRiT’s image quality comes at a high computational cost. Excessively long runtimes are a barrier to the clinical use of any reconstruction approach, and thus we discuss our approach to efficiently parallelizing ℓ1-SPIRiT and to achieving clinically-feasible runtimes. We present parallelizations of ℓ1-SPIRiT for both multi-GPU systems and multi-core CPUs, and discuss the software optimization and parallelization decisions made in our implementation. The performance of these alternatives depends on the processor architecture, the size of the image matrix, and the number of parallel imaging channels. Fundamentally, achieving fast runtime requires the correct trade-off between cache usage and parallelization overheads. We demonstrate image quality via a case from our clinical experimentation, using a custom 3DFT Spoiled Gradient Echo (SPGR) sequence with up to 8× acceleration via poisson-disc undersampling in the two phase-encoded directions. PMID:22345529

Parallel Multivariate Spatio-Temporal Clustering of Large Ecological Datasets on Hybrid Supercomputers

DOE Office of Scientific and Technical Information (OSTI.GOV)

Sreepathi, Sarat; Kumar, Jitendra; Mills, Richard T.

A proliferation of data from vast networks of remote sensing platforms (satellites, unmanned aircraft systems (UAS), airborne etc.), observational facilities (meteorological, eddy covariance etc.), state-of-the-art sensors, and simulation models offer unprecedented opportunities for scientific discovery. Unsupervised classification is a widely applied data mining approach to derive insights from such data. However, classification of very large data sets is a complex computational problem that requires efficient numerical algorithms and implementations on high performance computing (HPC) platforms. Additionally, increasing power, space, cooling and efficiency requirements has led to the deployment of hybrid supercomputing platforms with complex architectures and memory hierarchies like themore » Titan system at Oak Ridge National Laboratory. The advent of such accelerated computing architectures offers new challenges and opportunities for big data analytics in general and specifically, large scale cluster analysis in our case. Although there is an existing body of work on parallel cluster analysis, those approaches do not fully meet the needs imposed by the nature and size of our large data sets. Moreover, they had scaling limitations and were mostly limited to traditional distributed memory computing platforms. We present a parallel Multivariate Spatio-Temporal Clustering (MSTC) technique based on k-means cluster analysis that can target hybrid supercomputers like Titan. We developed a hybrid MPI, CUDA and OpenACC implementation that can utilize both CPU and GPU resources on computational nodes. We describe performance results on Titan that demonstrate the scalability and efficacy of our approach in processing large ecological data sets.« less

Scalable Static and Dynamic Community Detection Using Grappolo

DOE Office of Scientific and Technical Information (OSTI.GOV)

Halappanavar, Mahantesh; Lu, Hao; Kalyanaraman, Anantharaman

Graph clustering, popularly known as community detection, is a fundamental kernel for several applications of relevance to the Defense Advanced Research Projects Agency’s (DARPA) Hierarchical Identify Verify Exploit (HIVE) Pro- gram. Clusters or communities represent natural divisions within a network that are densely connected within a cluster and sparsely connected to the rest of the network. The need to compute clustering on large scale data necessitates the development of efficient algorithms that can exploit modern architectures that are fundamentally parallel in nature. How- ever, due to their irregular and inherently sequential nature, many of the current algorithms for community detectionmore » are challenging to parallelize. In response to the HIVE Graph Challenge, we present several parallelization heuristics for fast community detection using the Louvain method as the serial template. We implement all the heuristics in a software library called Grappolo. Using the inputs from the HIVE Challenge, we demonstrate superior performance and high quality solutions based on four parallelization heuristics. We use Grappolo on static graphs as the first step towards community detection on streaming graphs.« less

Multi-GPU parallel algorithm design and analysis for improved inversion of probability tomography with gravity gradiometry data

NASA Astrophysics Data System (ADS)

Hou, Zhenlong; Huang, Danian

2017-09-01

In this paper, we make a study on the inversion of probability tomography (IPT) with gravity gradiometry data at first. The space resolution of the results is improved by multi-tensor joint inversion, depth weighting matrix and the other methods. Aiming at solving the problems brought by the big data in the exploration, we present the parallel algorithm and the performance analysis combining Compute Unified Device Architecture (CUDA) with Open Multi-Processing (OpenMP) based on Graphics Processing Unit (GPU) accelerating. In the test of the synthetic model and real data from Vinton Dome, we get the improved results. It is also proved that the improved inversion algorithm is effective and feasible. The performance of parallel algorithm we designed is better than the other ones with CUDA. The maximum speedup could be more than 200. In the performance analysis, multi-GPU speedup and multi-GPU efficiency are applied to analyze the scalability of the multi-GPU programs. The designed parallel algorithm is demonstrated to be able to process larger scale of data and the new analysis method is practical.

On Parallelizing Single Dynamic Simulation Using HPC Techniques and APIs of Commercial Software

DOE Office of Scientific and Technical Information (OSTI.GOV)

Diao, Ruisheng; Jin, Shuangshuang; Howell, Frederic

Time-domain simulations are heavily used in today’s planning and operation practices to assess power system transient stability and post-transient voltage/frequency profiles following severe contingencies to comply with industry standards. Because of the increased modeling complexity, it is several times slower than real time for state-of-the-art commercial packages to complete a dynamic simulation for a large-scale model. With the growing stochastic behavior introduced by emerging technologies, power industry has seen a growing need for performing security assessment in real time. This paper presents a parallel implementation framework to speed up a single dynamic simulation by leveraging the existing stability model librarymore » in commercial tools through their application programming interfaces (APIs). Several high performance computing (HPC) techniques are explored such as parallelizing the calculation of generator current injection, identifying fast linear solvers for network solution, and parallelizing data outputs when interacting with APIs in the commercial package, TSAT. The proposed method has been tested on a WECC planning base case with detailed synchronous generator models and exhibits outstanding scalable performance with sufficient accuracy.« less

3D CSEM inversion based on goal-oriented adaptive finite element method

NASA Astrophysics Data System (ADS)

Zhang, Y.; Key, K.

2016-12-01

We present a parallel 3D frequency domain controlled-source electromagnetic inversion code name MARE3DEM. Non-linear inversion of observed data is performed with the Occam variant of regularized Gauss-Newton optimization. The forward operator is based on the goal-oriented finite element method that efficiently calculates the responses and sensitivity kernels in parallel using a data decomposition scheme where independent modeling tasks contain different frequencies and subsets of the transmitters and receivers. To accommodate complex 3D conductivity variation with high flexibility and precision, we adopt the dual-grid approach where the forward mesh conforms to the inversion parameter grid and is adaptively refined until the forward solution converges to the desired accuracy. This dual-grid approach is memory efficient, since the inverse parameter grid remains independent from fine meshing generated around the transmitter and receivers by the adaptive finite element method. Besides, the unstructured inverse mesh efficiently handles multiple scale structures and allows for fine-scale model parameters within the region of interest. Our mesh generation engine keeps track of the refinement hierarchy so that the map of conductivity and sensitivity kernel between the forward and inverse mesh is retained. We employ the adjoint-reciprocity method to calculate the sensitivity kernels which establish a linear relationship between changes in the conductivity model and changes in the modeled responses. Our code uses a direcy solver for the linear systems, so the adjoint problem is efficiently computed by re-using the factorization from the primary problem. Further computational efficiency and scalability is obtained in the regularized Gauss-Newton portion of the inversion using parallel dense matrix-matrix multiplication and matrix factorization routines implemented with the ScaLAPACK library. We show the scalability, reliability and the potential of the algorithm to deal with complex geological scenarios by applying it to the inversion of synthetic marine controlled source EM data generated for a complex 3D offshore model with significant seafloor topography.

2D Seismic Imaging of Elastic Parameters by Frequency Domain Full Waveform Inversion

NASA Astrophysics Data System (ADS)

Brossier, R.; Virieux, J.; Operto, S.

2008-12-01

Thanks to recent advances in parallel computing, full waveform inversion is today a tractable seismic imaging method to reconstruct physical parameters of the earth interior at different scales ranging from the near- surface to the deep crust. We present a massively parallel 2D frequency-domain full-waveform algorithm for imaging visco-elastic media from multi-component seismic data. The forward problem (i.e. the resolution of the frequency-domain 2D PSV elastodynamics equations) is based on low-order Discontinuous Galerkin (DG) method (P0 and/or P1 interpolations). Thanks to triangular unstructured meshes, the DG method allows accurate modeling of both body waves and surface waves in case of complex topography for a discretization of 10 to 15 cells per shear wavelength. The frequency-domain DG system is solved efficiently for multiple sources with the parallel direct solver MUMPS. The local inversion procedure (i.e. minimization of residuals between observed and computed data) is based on the adjoint-state method which allows to efficiently compute the gradient of the objective function. Applying the inversion hierarchically from the low frequencies to the higher ones defines a multiresolution imaging strategy which helps convergence towards the global minimum. In place of expensive Newton algorithm, the combined use of the diagonal terms of the approximate Hessian matrix and optimization algorithms based on quasi-Newton methods (Conjugate Gradient, LBFGS, ...) allows to improve the convergence of the iterative inversion. The distribution of forward problem solutions over processors driven by a mesh partitioning performed by METIS allows to apply most of the inversion in parallel. We shall present the main features of the parallel modeling/inversion algorithm, assess its scalability and illustrate its performances with realistic synthetic case studies.

Development and Validation of a Fast, Accurate and Cost-Effective Aeroservoelastic Method on Advanced Parallel Computing Systems

NASA Technical Reports Server (NTRS)

Goodwin, Sabine A.; Raj, P.

1999-01-01

Progress to date towards the development and validation of a fast, accurate and cost-effective aeroelastic method for advanced parallel computing platforms such as the IBM SP2 and the SGI Origin 2000 is presented in this paper. The ENSAERO code, developed at the NASA-Ames Research Center has been selected for this effort. The code allows for the computation of aeroelastic responses by simultaneously integrating the Euler or Navier-Stokes equations and the modal structural equations of motion. To assess the computational performance and accuracy of the ENSAERO code, this paper reports the results of the Navier-Stokes simulations of the transonic flow over a flexible aeroelastic wing body configuration. In addition, a forced harmonic oscillation analysis in the frequency domain and an analysis in the time domain are done on a wing undergoing a rigid pitch and plunge motion. Finally, to demonstrate the ENSAERO flutter-analysis capability, aeroelastic Euler and Navier-Stokes computations on an L-1011 wind tunnel model including pylon, nacelle and empennage are underway. All computational solutions are compared with experimental data to assess the level of accuracy of ENSAERO. As the computations described above are performed, a meticulous log of computational performance in terms of wall clock time, execution speed, memory and disk storage is kept. Code scalability is also demonstrated by studying the impact of varying the number of processors on computational performance on the IBM SP2 and the Origin 2000 systems.

Climate Data Assimilation on a Massively Parallel Supercomputer

NASA Technical Reports Server (NTRS)

Ding, Hong Q.; Ferraro, Robert D.

1996-01-01

We have designed and implemented a set of highly efficient and highly scalable algorithms for an unstructured computational package, the PSAS data assimilation package, as demonstrated by detailed performance analysis of systematic runs on up to 512-nodes of an Intel Paragon. The preconditioned Conjugate Gradient solver achieves a sustained 18 Gflops performance. Consequently, we achieve an unprecedented 100-fold reduction in time to solution on the Intel Paragon over a single head of a Cray C90. This not only exceeds the daily performance requirement of the Data Assimilation Office at NASA's Goddard Space Flight Center, but also makes it possible to explore much larger and challenging data assimilation problems which are unthinkable on a traditional computer platform such as the Cray C90.

Multi-Purpose, Application-Centric, Scalable I/O Proxy Application

DOE Office of Scientific and Technical Information (OSTI.GOV)

Miller, M. C.

2015-06-15

MACSio is a Multi-purpose, Application-Centric, Scalable I/O proxy application. It is designed to support a number of goals with respect to parallel I/O performance testing and benchmarking including the ability to test and compare various I/O libraries and I/O paradigms, to predict scalable performance of real applications and to help identify where improvements in I/O performance can be made within the HPC I/O software stack.

The Automated Instrumentation and Monitoring System (AIMS): Design and Architecture. 3.2

NASA Technical Reports Server (NTRS)

Yan, Jerry C.; Schmidt, Melisa; Schulbach, Cathy; Bailey, David (Technical Monitor)

1997-01-01

Whether a researcher is designing the 'next parallel programming paradigm', another 'scalable multiprocessor' or investigating resource allocation algorithms for multiprocessors, a facility that enables parallel program execution to be captured and displayed is invaluable. Careful analysis of such information can help computer and software architects to capture, and therefore, exploit behavioral variations among/within various parallel programs to take advantage of specific hardware characteristics. A software tool-set that facilitates performance evaluation of parallel applications on multiprocessors has been put together at NASA Ames Research Center under the sponsorship of NASA's High Performance Computing and Communications Program over the past five years. The Automated Instrumentation and Monitoring Systematic has three major software components: a source code instrumentor which automatically inserts active event recorders into program source code before compilation; a run-time performance monitoring library which collects performance data; and a visualization tool-set which reconstructs program execution based on the data collected. Besides being used as a prototype for developing new techniques for instrumenting, monitoring and presenting parallel program execution, AIMS is also being incorporated into the run-time environments of various hardware testbeds to evaluate their impact on user productivity. Currently, the execution of FORTRAN and C programs on the Intel Paragon and PALM workstations can be automatically instrumented and monitored. Performance data thus collected can be displayed graphically on various workstations. The process of performance tuning with AIMS will be illustrated using various NAB Parallel Benchmarks. This report includes a description of the internal architecture of AIMS and a listing of the source code.

A divide-conquer-recombine algorithmic paradigm for large spatiotemporal quantum molecular dynamics simulations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Shimojo, Fuyuki; Hattori, Shinnosuke; Department of Physics, Kumamoto University, Kumamoto 860-8555

We introduce an extension of the divide-and-conquer (DC) algorithmic paradigm called divide-conquer-recombine (DCR) to perform large quantum molecular dynamics (QMD) simulations on massively parallel supercomputers, in which interatomic forces are computed quantum mechanically in the framework of density functional theory (DFT). In DCR, the DC phase constructs globally informed, overlapping local-domain solutions, which in the recombine phase are synthesized into a global solution encompassing large spatiotemporal scales. For the DC phase, we design a lean divide-and-conquer (LDC) DFT algorithm, which significantly reduces the prefactor of the O(N) computational cost for N electrons by applying a density-adaptive boundary condition at themore » peripheries of the DC domains. Our globally scalable and locally efficient solver is based on a hybrid real-reciprocal space approach that combines: (1) a highly scalable real-space multigrid to represent the global charge density; and (2) a numerically efficient plane-wave basis for local electronic wave functions and charge density within each domain. Hybrid space-band decomposition is used to implement the LDC-DFT algorithm on parallel computers. A benchmark test on an IBM Blue Gene/Q computer exhibits an isogranular parallel efficiency of 0.984 on 786 432 cores for a 50.3 × 10{sup 6}-atom SiC system. As a test of production runs, LDC-DFT-based QMD simulation involving 16 661 atoms is performed on the Blue Gene/Q to study on-demand production of hydrogen gas from water using LiAl alloy particles. As an example of the recombine phase, LDC-DFT electronic structures are used as a basis set to describe global photoexcitation dynamics with nonadiabatic QMD (NAQMD) and kinetic Monte Carlo (KMC) methods. The NAQMD simulations are based on the linear response time-dependent density functional theory to describe electronic excited states and a surface-hopping approach to describe transitions between the excited states. A series of techniques are employed for efficiently calculating the long-range exact exchange correction and excited-state forces. The NAQMD trajectories are analyzed to extract the rates of various excitonic processes, which are then used in KMC simulation to study the dynamics of the global exciton flow network. This has allowed the study of large-scale photoexcitation dynamics in 6400-atom amorphous molecular solid, reaching the experimental time scales.« less

Space Situational Awareness Data Processing Scalability Utilizing Google Cloud Services

NASA Astrophysics Data System (ADS)

Greenly, D.; Duncan, M.; Wysack, J.; Flores, F.

Space Situational Awareness (SSA) is a fundamental and critical component of current space operations. The term SSA encompasses the awareness, understanding and predictability of all objects in space. As the population of orbital space objects and debris increases, the number of collision avoidance maneuvers grows and prompts the need for accurate and timely process measures. The SSA mission continually evolves to near real-time assessment and analysis demanding the need for higher processing capabilities. By conventional methods, meeting these demands requires the integration of new hardware to keep pace with the growing complexity of maneuver planning algorithms. SpaceNav has implemented a highly scalable architecture that will track satellites and debris by utilizing powerful virtual machines on the Google Cloud Platform. SpaceNav algorithms for processing CDMs outpace conventional means. A robust processing environment for tracking data, collision avoidance maneuvers and various other aspects of SSA can be created and deleted on demand. Migrating SpaceNav tools and algorithms into the Google Cloud Platform will be discussed and the trials and tribulations involved. Information will be shared on how and why certain cloud products were used as well as integration techniques that were implemented. Key items to be presented are: 1.Scientific algorithms and SpaceNav tools integrated into a scalable architecture a) Maneuver Planning b) Parallel Processing c) Monte Carlo Simulations d) Optimization Algorithms e) SW Application Development/Integration into the Google Cloud Platform 2. Compute Engine Processing a) Application Engine Automated Processing b) Performance testing and Performance Scalability c) Cloud MySQL databases and Database Scalability d) Cloud Data Storage e) Redundancy and Availability

GATECloud.net: a platform for large-scale, open-source text processing on the cloud.

PubMed

Tablan, Valentin; Roberts, Ian; Cunningham, Hamish; Bontcheva, Kalina

2013-01-28

Cloud computing is increasingly being regarded as a key enabler of the 'democratization of science', because on-demand, highly scalable cloud computing facilities enable researchers anywhere to carry out data-intensive experiments. In the context of natural language processing (NLP), algorithms tend to be complex, which makes their parallelization and deployment on cloud platforms a non-trivial task. This study presents a new, unique, cloud-based platform for large-scale NLP research--GATECloud. net. It enables researchers to carry out data-intensive NLP experiments by harnessing the vast, on-demand compute power of the Amazon cloud. Important infrastructural issues are dealt with by the platform, completely transparently for the researcher: load balancing, efficient data upload and storage, deployment on the virtual machines, security and fault tolerance. We also include a cost-benefit analysis and usage evaluation.

Using NCAR Yellowstone for PhotoVoltaic Power Forecasts with Artificial Neural Networks and an Analog Ensemble

NASA Astrophysics Data System (ADS)

Cervone, G.; Clemente-Harding, L.; Alessandrini, S.; Delle Monache, L.

2016-12-01

A methodology based on Artificial Neural Networks (ANN) and an Analog Ensemble (AnEn) is presented to generate 72-hour deterministic and probabilistic forecasts of power generated by photovoltaic (PV) power plants using input from a numerical weather prediction model and computed astronomical variables. ANN and AnEn are used individually and in combination to generate forecasts for three solar power plant located in Italy. The computational scalability of the proposed solution is tested using synthetic data simulating 4,450 PV power stations. The NCAR Yellowstone supercomputer is employed to test the parallel implementation of the proposed solution, ranging from 1 node (32 cores) to 4,450 nodes (141,140 cores). Results show that a combined AnEn + ANN solution yields best results, and that the proposed solution is well suited for massive scale computation.

Scalable Molecular Dynamics with NAMD

PubMed Central

Phillips, James C.; Braun, Rosemary; Wang, Wei; Gumbart, James; Tajkhorshid, Emad; Villa, Elizabeth; Chipot, Christophe; Skeel, Robert D.; Kalé, Laxmikant; Schulten, Klaus

2008-01-01

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD scales to hundreds of processors on high-end parallel platforms, as well as tens of processors on low-cost commodity clusters, and also runs on individual desktop and laptop computers. NAMD works with AMBER and CHARMM potential functions, parameters, and file formats. This paper, directed to novices as well as experts, first introduces concepts and methods used in the NAMD program, describing the classical molecular dynamics force field, equations of motion, and integration methods along with the efficient electrostatics evaluation algorithms employed and temperature and pressure controls used. Features for steering the simulation across barriers and for calculating both alchemical and conformational free energy differences are presented. The motivations for and a roadmap to the internal design of NAMD, implemented in C++ and based on Charm++ parallel objects, are outlined. The factors affecting the serial and parallel performance of a simulation are discussed. Next, typical NAMD use is illustrated with representative applications to a small, a medium, and a large biomolecular system, highlighting particular features of NAMD, e.g., the Tcl scripting language. Finally, the paper provides a list of the key features of NAMD and discusses the benefits of combining NAMD with the molecular graphics/sequence analysis software VMD and the grid computing/collaboratory software BioCoRE. NAMD is distributed free of charge with source code at www.ks.uiuc.edu. PMID:16222654

GROMACS 4:  Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation.

PubMed

Hess, Berk; Kutzner, Carsten; van der Spoel, David; Lindahl, Erik

2008-03-01

Molecular simulation is an extremely useful, but computationally very expensive tool for studies of chemical and biomolecular systems. Here, we present a new implementation of our molecular simulation toolkit GROMACS which now both achieves extremely high performance on single processors from algorithmic optimizations and hand-coded routines and simultaneously scales very well on parallel machines. The code encompasses a minimal-communication domain decomposition algorithm, full dynamic load balancing, a state-of-the-art parallel constraint solver, and efficient virtual site algorithms that allow removal of hydrogen atom degrees of freedom to enable integration time steps up to 5 fs for atomistic simulations also in parallel. To improve the scaling properties of the common particle mesh Ewald electrostatics algorithms, we have in addition used a Multiple-Program, Multiple-Data approach, with separate node domains responsible for direct and reciprocal space interactions. Not only does this combination of algorithms enable extremely long simulations of large systems but also it provides that simulation performance on quite modest numbers of standard cluster nodes.

The Software Correlator of the Chinese VLBI Network

NASA Technical Reports Server (NTRS)

Zheng, Weimin; Quan, Ying; Shu, Fengchun; Chen, Zhong; Chen, Shanshan; Wang, Weihua; Wang, Guangli

2010-01-01

The software correlator of the Chinese VLBI Network (CVN) has played an irreplaceable role in the CVN routine data processing, e.g., in the Chinese lunar exploration project. This correlator will be upgraded to process geodetic and astronomical observation data. In the future, with several new stations joining the network, CVN will carry out crustal movement observations, quick UT1 measurements, astrophysical observations, and deep space exploration activities. For the geodetic or astronomical observations, we need a wide-band 10-station correlator. For spacecraft tracking, a realtime and highly reliable correlator is essential. To meet the scientific and navigation requirements of CVN, two parallel software correlators in the multiprocessor environments are under development. A high speed, 10-station prototype correlator using the mixed Pthreads and MPI (Massage Passing Interface) parallel algorithm on a computer cluster platform is being developed. Another real-time software correlator for spacecraft tracking adopts the thread-parallel technology, and it runs on the SMP (Symmetric Multiple Processor) servers. Both correlators have the characteristic of flexible structure and scalability.

Extreme-Scale Bayesian Inference for Uncertainty Quantification of Complex Simulations

DOE Office of Scientific and Technical Information (OSTI.GOV)

Biros, George

Uncertainty quantification (UQ)—that is, quantifying uncertainties in complex mathematical models and their large-scale computational implementations—is widely viewed as one of the outstanding challenges facing the field of CS&E over the coming decade. The EUREKA project set to address the most difficult class of UQ problems: those for which both the underlying PDE model as well as the uncertain parameters are of extreme scale. In the project we worked on these extreme-scale challenges in the following four areas: 1. Scalable parallel algorithms for sampling and characterizing the posterior distribution that exploit the structure of the underlying PDEs and parameter-to-observable map. Thesemore » include structure-exploiting versions of the randomized maximum likelihood method, which aims to overcome the intractability of employing conventional MCMC methods for solving extreme-scale Bayesian inversion problems by appealing to and adapting ideas from large-scale PDE-constrained optimization, which have been very successful at exploring high-dimensional spaces. 2. Scalable parallel algorithms for construction of prior and likelihood functions based on learning methods and non-parametric density estimation. Constructing problem-specific priors remains a critical challenge in Bayesian inference, and more so in high dimensions. Another challenge is construction of likelihood functions that capture unmodeled couplings between observations and parameters. We will create parallel algorithms for non-parametric density estimation using high dimensional N-body methods and combine them with supervised learning techniques for the construction of priors and likelihood functions. 3. Bayesian inadequacy models, which augment physics models with stochastic models that represent their imperfections. The success of the Bayesian inference framework depends on the ability to represent the uncertainty due to imperfections of the mathematical model of the phenomena of interest. This is a central challenge in UQ, especially for large-scale models. We propose to develop the mathematical tools to address these challenges in the context of extreme-scale problems. 4. Parallel scalable algorithms for Bayesian optimal experimental design (OED). Bayesian inversion yields quantified uncertainties in the model parameters, which can be propagated forward through the model to yield uncertainty in outputs of interest. This opens the way for designing new experiments to reduce the uncertainties in the model parameters and model predictions. Such experimental design problems have been intractable for large-scale problems using conventional methods; we will create OED algorithms that exploit the structure of the PDE model and the parameter-to-output map to overcome these challenges. Parallel algorithms for these four problems were created, analyzed, prototyped, implemented, tuned, and scaled up for leading-edge supercomputers, including UT-Austin’s own 10 petaflops Stampede system, ANL’s Mira system, and ORNL’s Titan system. While our focus is on fundamental mathematical/computational methods and algorithms, we will assess our methods on model problems derived from several DOE mission applications, including multiscale mechanics and ice sheet dynamics.« less

Slices: A Scalable Partitioner for Finite Element Meshes

NASA Technical Reports Server (NTRS)

Ding, H. Q.; Ferraro, R. D.

1995-01-01

A parallel partitioner for partitioning unstructured finite element meshes on distributed memory architectures is developed. The element based partitioner can handle mixtures of different element types. All algorithms adopted in the partitioner are scalable, including a communication template for unpredictable incoming messages, as shown in actual timing measurements.

Scalable and fast heterogeneous molecular simulation with predictive parallelization schemes

DOE Office of Scientific and Technical Information (OSTI.GOV)

Guzman, Horacio V.; Junghans, Christoph; Kremer, Kurt

Multiscale and inhomogeneous molecular systems are challenging topics in the field of molecular simulation. In particular, modeling biological systems in the context of multiscale simulations and exploring material properties are driving a permanent development of new simulation methods and optimization algorithms. In computational terms, those methods require parallelization schemes that make a productive use of computational resources for each simulation and from its genesis. Here, we introduce the heterogeneous domain decomposition approach, which is a combination of an heterogeneity-sensitive spatial domain decomposition with an a priori rearrangement of subdomain walls. Within this approach and paper, the theoretical modeling and scalingmore » laws for the force computation time are proposed and studied as a function of the number of particles and the spatial resolution ratio. We also show the new approach capabilities, by comparing it to both static domain decomposition algorithms and dynamic load-balancing schemes. Specifically, two representative molecular systems have been simulated and compared to the heterogeneous domain decomposition proposed in this work. Finally, these two systems comprise an adaptive resolution simulation of a biomolecule solvated in water and a phase-separated binary Lennard-Jones fluid.« less

Scalable and fast heterogeneous molecular simulation with predictive parallelization schemes

DOE PAGES

Guzman, Horacio V.; Junghans, Christoph; Kremer, Kurt; ...

2017-11-27

Multiscale and inhomogeneous molecular systems are challenging topics in the field of molecular simulation. In particular, modeling biological systems in the context of multiscale simulations and exploring material properties are driving a permanent development of new simulation methods and optimization algorithms. In computational terms, those methods require parallelization schemes that make a productive use of computational resources for each simulation and from its genesis. Here, we introduce the heterogeneous domain decomposition approach, which is a combination of an heterogeneity-sensitive spatial domain decomposition with an a priori rearrangement of subdomain walls. Within this approach and paper, the theoretical modeling and scalingmore » laws for the force computation time are proposed and studied as a function of the number of particles and the spatial resolution ratio. We also show the new approach capabilities, by comparing it to both static domain decomposition algorithms and dynamic load-balancing schemes. Specifically, two representative molecular systems have been simulated and compared to the heterogeneous domain decomposition proposed in this work. Finally, these two systems comprise an adaptive resolution simulation of a biomolecule solvated in water and a phase-separated binary Lennard-Jones fluid.« less

(Re)engineering Earth System Models to Expose Greater Concurrency for Ultrascale Computing: Practice, Experience, and Musings

NASA Astrophysics Data System (ADS)

Mills, R. T.

2014-12-01

As the high performance computing (HPC) community pushes towards the exascale horizon, the importance and prevalence of fine-grained parallelism in new computer architectures is increasing. This is perhaps most apparent in the proliferation of so-called "accelerators" such as the Intel Xeon Phi or NVIDIA GPGPUs, but the trend also holds for CPUs, where serial performance has grown slowly and effective use of hardware threads and vector units are becoming increasingly important to realizing high performance. This has significant implications for weather, climate, and Earth system modeling codes, many of which display impressive scalability across MPI ranks but take relatively little advantage of threading and vector processing. In addition to increasing parallelism, next generation codes will also need to address increasingly deep hierarchies for data movement: NUMA/cache levels, on node vs. off node, local vs. wide neighborhoods on the interconnect, and even in the I/O system. We will discuss some approaches (grounded in experiences with the Intel Xeon Phi architecture) for restructuring Earth science codes to maximize concurrency across multiple levels (vectors, threads, MPI ranks), and also discuss some novel approaches for minimizing expensive data movement/communication.

Simulation of 2D Kinetic Effects in Plasmas using the Grid Based Continuum Code LOKI

NASA Astrophysics Data System (ADS)

Banks, Jeffrey; Berger, Richard; Chapman, Tom; Brunner, Stephan

2016-10-01

Kinetic simulation of multi-dimensional plasma waves through direct discretization of the Vlasov equation is a useful tool to study many physical interactions and is particularly attractive for situations where minimal fluctuation levels are desired, for instance, when measuring growth rates of plasma wave instabilities. However, direct discretization of phase space can be computationally expensive, and as a result there are few examples of published results using Vlasov codes in more than a single configuration space dimension. In an effort to fill this gap we have developed the Eulerian-based kinetic code LOKI that evolves the Vlasov-Poisson system in 2+2-dimensional phase space. The code is designed to reduce the cost of phase-space computation by using fully 4th order accurate conservative finite differencing, while retaining excellent parallel scalability that efficiently uses large scale computing resources. In this poster I will discuss the algorithms used in the code as well as some aspects of their parallel implementation using MPI. I will also overview simulation results of basic plasma wave instabilities relevant to laser plasma interaction, which have been obtained using the code.

Highlights of X-Stack ExM Deliverable Swift/T

DOE Office of Scientific and Technical Information (OSTI.GOV)

Wozniak, Justin M.

Swift/T is a key success from the ExM: System support for extreme-scale, many-task applications1 X-Stack project, which proposed to use concurrent dataflow as an innovative programming model to exploit extreme parallelism in exascale computers. The Swift/T component of the project reimplemented the Swift language from scratch to allow applications that compose scientific modules together to be build and run on available petascale computers (Blue Gene, Cray). Swift/T does this via a new compiler and runtime that generates and executes the application as an MPI program. We assume that mission-critical emerging exascale applications will be composed as scalable applications using existingmore » software components, connected by data dependencies. Developers wrap native code fragments using a higherlevel language, then build composite applications to form a computational experiment. This exemplifies hierarchical concurrency: lower-level messaging libraries are used for fine-grained parallelism; highlevel control is used for inter-task coordination. These patterns are best expressed with dataflow, but static DAGs (i.e., other workflow languages) limit the applications that can be built; they do not provide the expressiveness of Swift, such as conditional execution, iteration, and recursive functions.« less

Parallel multiscale simulations of a brain aneurysm

PubMed Central

Grinberg, Leopold; Fedosov, Dmitry A.; Karniadakis, George Em

2012-01-01

Cardiovascular pathologies, such as a brain aneurysm, are affected by the global blood circulation as well as by the local microrheology. Hence, developing computational models for such cases requires the coupling of disparate spatial and temporal scales often governed by diverse mathematical descriptions, e.g., by partial differential equations (continuum) and ordinary differential equations for discrete particles (atomistic). However, interfacing atomistic-based with continuum-based domain discretizations is a challenging problem that requires both mathematical and computational advances. We present here a hybrid methodology that enabled us to perform the first multi-scale simulations of platelet depositions on the wall of a brain aneurysm. The large scale flow features in the intracranial network are accurately resolved by using the high-order spectral element Navier-Stokes solver εκ αr. The blood rheology inside the aneurysm is modeled using a coarse-grained stochastic molecular dynamics approach (the dissipative particle dynamics method) implemented in the parallel code LAMMPS. The continuum and atomistic domains overlap with interface conditions provided by effective forces computed adaptively to ensure continuity of states across the interface boundary. A two-way interaction is allowed with the time-evolving boundary of the (deposited) platelet clusters tracked by an immersed boundary method. The corresponding heterogeneous solvers ( εκ αr and LAMMPS) are linked together by a computational multilevel message passing interface that facilitates modularity and high parallel efficiency. Results of multiscale simulations of clot formation inside the aneurysm in a patient-specific arterial tree are presented. We also discuss the computational challenges involved and present scalability results of our coupled solver on up to 300K computer processors. Validation of such coupled atomistic-continuum models is a main open issue that has to be addressed in future work. PMID:23734066

Parallel computing in enterprise modeling.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Goldsby, Michael E.; Armstrong, Robert C.; Shneider, Max S.

2008-08-01

This report presents the results of our efforts to apply high-performance computing to entity-based simulations with a multi-use plugin for parallel computing. We use the term 'Entity-based simulation' to describe a class of simulation which includes both discrete event simulation and agent based simulation. What simulations of this class share, and what differs from more traditional models, is that the result sought is emergent from a large number of contributing entities. Logistic, economic and social simulations are members of this class where things or people are organized or self-organize to produce a solution. Entity-based problems never have an a priorimore » ergodic principle that will greatly simplify calculations. Because the results of entity-based simulations can only be realized at scale, scalable computing is de rigueur for large problems. Having said that, the absence of a spatial organizing principal makes the decomposition of the problem onto processors problematic. In addition, practitioners in this domain commonly use the Java programming language which presents its own problems in a high-performance setting. The plugin we have developed, called the Parallel Particle Data Model, overcomes both of these obstacles and is now being used by two Sandia frameworks: the Decision Analysis Center, and the Seldon social simulation facility. While the ability to engage U.S.-sized problems is now available to the Decision Analysis Center, this plugin is central to the success of Seldon. Because Seldon relies on computationally intensive cognitive sub-models, this work is necessary to achieve the scale necessary for realistic results. With the recent upheavals in the financial markets, and the inscrutability of terrorist activity, this simulation domain will likely need a capability with ever greater fidelity. High-performance computing will play an important part in enabling that greater fidelity.« less

Parallel multiscale simulations of a brain aneurysm.

PubMed

Grinberg, Leopold; Fedosov, Dmitry A; Karniadakis, George Em

2013-07-01

Cardiovascular pathologies, such as a brain aneurysm, are affected by the global blood circulation as well as by the local microrheology. Hence, developing computational models for such cases requires the coupling of disparate spatial and temporal scales often governed by diverse mathematical descriptions, e.g., by partial differential equations (continuum) and ordinary differential equations for discrete particles (atomistic). However, interfacing atomistic-based with continuum-based domain discretizations is a challenging problem that requires both mathematical and computational advances. We present here a hybrid methodology that enabled us to perform the first multi-scale simulations of platelet depositions on the wall of a brain aneurysm. The large scale flow features in the intracranial network are accurately resolved by using the high-order spectral element Navier-Stokes solver εκ αr . The blood rheology inside the aneurysm is modeled using a coarse-grained stochastic molecular dynamics approach (the dissipative particle dynamics method) implemented in the parallel code LAMMPS. The continuum and atomistic domains overlap with interface conditions provided by effective forces computed adaptively to ensure continuity of states across the interface boundary. A two-way interaction is allowed with the time-evolving boundary of the (deposited) platelet clusters tracked by an immersed boundary method. The corresponding heterogeneous solvers ( εκ αr and LAMMPS) are linked together by a computational multilevel message passing interface that facilitates modularity and high parallel efficiency. Results of multiscale simulations of clot formation inside the aneurysm in a patient-specific arterial tree are presented. We also discuss the computational challenges involved and present scalability results of our coupled solver on up to 300K computer processors. Validation of such coupled atomistic-continuum models is a main open issue that has to be addressed in future work.

Parallel multiscale simulations of a brain aneurysm

DOE Office of Scientific and Technical Information (OSTI.GOV)

Grinberg, Leopold; Fedosov, Dmitry A.; Karniadakis, George Em, E-mail: george_karniadakis@brown.edu

2013-07-01

Cardiovascular pathologies, such as a brain aneurysm, are affected by the global blood circulation as well as by the local microrheology. Hence, developing computational models for such cases requires the coupling of disparate spatial and temporal scales often governed by diverse mathematical descriptions, e.g., by partial differential equations (continuum) and ordinary differential equations for discrete particles (atomistic). However, interfacing atomistic-based with continuum-based domain discretizations is a challenging problem that requires both mathematical and computational advances. We present here a hybrid methodology that enabled us to perform the first multiscale simulations of platelet depositions on the wall of a brain aneurysm.more » The large scale flow features in the intracranial network are accurately resolved by using the high-order spectral element Navier–Stokes solver NεκTαr. The blood rheology inside the aneurysm is modeled using a coarse-grained stochastic molecular dynamics approach (the dissipative particle dynamics method) implemented in the parallel code LAMMPS. The continuum and atomistic domains overlap with interface conditions provided by effective forces computed adaptively to ensure continuity of states across the interface boundary. A two-way interaction is allowed with the time-evolving boundary of the (deposited) platelet clusters tracked by an immersed boundary method. The corresponding heterogeneous solvers (NεκTαr and LAMMPS) are linked together by a computational multilevel message passing interface that facilitates modularity and high parallel efficiency. Results of multiscale simulations of clot formation inside the aneurysm in a patient-specific arterial tree are presented. We also discuss the computational challenges involved and present scalability results of our coupled solver on up to 300 K computer processors. Validation of such coupled atomistic-continuum models is a main open issue that has to be addressed in future work.« less

An approach to computing discrete adjoints for MPI-parallelized models applied to Ice Sheet System Model 4.11

NASA Astrophysics Data System (ADS)

Larour, Eric; Utke, Jean; Bovin, Anton; Morlighem, Mathieu; Perez, Gilberto

2016-11-01

Within the framework of sea-level rise projections, there is a strong need for hindcast validation of the evolution of polar ice sheets in a way that tightly matches observational records (from radar, gravity, and altimetry observations mainly). However, the computational requirements for making hindcast reconstructions possible are severe and rely mainly on the evaluation of the adjoint state of transient ice-flow models. Here, we look at the computation of adjoints in the context of the NASA/JPL/UCI Ice Sheet System Model (ISSM), written in C++ and designed for parallel execution with MPI. We present the adaptations required in the way the software is designed and written, but also generic adaptations in the tools facilitating the adjoint computations. We concentrate on the use of operator overloading coupled with the AdjoinableMPI library to achieve the adjoint computation of the ISSM. We present a comprehensive approach to (1) carry out type changing through the ISSM, hence facilitating operator overloading, (2) bind to external solvers such as MUMPS and GSL-LU, and (3) handle MPI-based parallelism to scale the capability. We demonstrate the success of the approach by computing sensitivities of hindcast metrics such as the misfit to observed records of surface altimetry on the northeastern Greenland Ice Stream, or the misfit to observed records of surface velocities on Upernavik Glacier, central West Greenland. We also provide metrics for the scalability of the approach, and the expected performance. This approach has the potential to enable a new generation of hindcast-validated projections that make full use of the wealth of datasets currently being collected, or already collected, in Greenland and Antarctica.

An Approach to Computing Discrete Adjoints for MPI-Parallelized Models Applied to the Ice Sheet System Model}

NASA Astrophysics Data System (ADS)

Perez, G. L.; Larour, E. Y.; Morlighem, M.

2016-12-01

Within the framework of sea-level rise projections, there is a strong need for hindcast validation of the evolution of polar ice sheets in a way that tightly matches observational records (from radar and altimetry observations mainly). However, the computational requirements for making hindcast reconstructions possible are severe and rely mainly on the evaluation of the adjoint state of transient ice-flow models. Here, we look at the computation of adjoints in the context of the NASA/JPL/UCI Ice Sheet System Model, written in C++ and designed for parallel execution with MPI. We present the adaptations required in the way the software is designed and written but also generic adaptations in the tools facilitating the adjoint computations. We concentrate on the use of operator overloading coupled with the AdjoinableMPI library to achieve the adjoint computation of ISSM. We present a comprehensive approach to 1) carry out type changing through ISSM, hence facilitating operator overloading, 2) bind to external solvers such as MUMPS and GSL-LU and 3) handle MPI-based parallelism to scale the capability. We demonstrate the success of the approach by computing sensitivities of hindcast metrics such as the misfit to observed records of surface altimetry on the North-East Greenland Ice Stream, or the misfit to observed records of surface velocities on Upernavik Glacier, Central West Greenland. We also provide metrics for the scalability of the approach, and the expected performance. This approach has the potential of enabling a new generation of hindcast-validated projections that make full use of the wealth of datasets currently being collected, or alreay collected in Greenland and Antarctica, such as surface altimetry, surface velocities, and/or gravity measurements.

Advanced Technology and Mitigation (ATDM) SPARC Re-Entry Code Fiscal Year 2017 Progress and Accomplishments for ECP.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Crozier, Paul; Howard, Micah; Rider, William J.

The SPARC (Sandia Parallel Aerodynamics and Reentry Code) will provide nuclear weapon qualification evidence for the random vibration and thermal environments created by re-entry of a warhead into the earth’s atmosphere. SPARC incorporates the innovative approaches of ATDM projects on several fronts including: effective harnessing of heterogeneous compute nodes using Kokkos, exascale-ready parallel scalability through asynchronous multi-tasking, uncertainty quantification through Sacado integration, implementation of state-of-the-art reentry physics and multiscale models, use of advanced verification and validation methods, and enabling of improved workflows for users. SPARC is being developed primarily for the Department of Energy nuclear weapon program, with additional developmentmore » and use of the code is being supported by the Department of Defense for conventional weapons programs.« less

ICON-MIC: Implementing a CPU/MIC Collaboration Parallel Framework for ICON on Tianhe-2 Supercomputer.

PubMed

Wang, Zihao; Chen, Yu; Zhang, Jingrong; Li, Lun; Wan, Xiaohua; Liu, Zhiyong; Sun, Fei; Zhang, Fa

2018-03-01

Electron tomography (ET) is an important technique for studying the three-dimensional structures of the biological ultrastructure. Recently, ET has reached sub-nanometer resolution for investigating the native and conformational dynamics of macromolecular complexes by combining with the sub-tomogram averaging approach. Due to the limited sampling angles, ET reconstruction typically suffers from the "missing wedge" problem. Using a validation procedure, iterative compressed-sensing optimized nonuniform fast Fourier transform (NUFFT) reconstruction (ICON) demonstrates its power in restoring validated missing information for a low-signal-to-noise ratio biological ET dataset. However, the huge computational demand has become a bottleneck for the application of ICON. In this work, we implemented a parallel acceleration technology ICON-many integrated core (MIC) on Xeon Phi cards to address the huge computational demand of ICON. During this step, we parallelize the element-wise matrix operations and use the efficient summation of a matrix to reduce the cost of matrix computation. We also developed parallel versions of NUFFT on MIC to achieve a high acceleration of ICON by using more efficient fast Fourier transform (FFT) calculation. We then proposed a hybrid task allocation strategy (two-level load balancing) to improve the overall performance of ICON-MIC by making full use of the idle resources on Tianhe-2 supercomputer. Experimental results using two different datasets show that ICON-MIC has high accuracy in biological specimens under different noise levels and a significant acceleration, up to 13.3 × , compared with the CPU version. Further, ICON-MIC has good scalability efficiency and overall performance on Tianhe-2 supercomputer.

High-Speed Computation of the Kleene Star in Max-Plus Algebraic System Using a Cell Broadband Engine

NASA Astrophysics Data System (ADS)

Goto, Hiroyuki

This research addresses a high-speed computation method for the Kleene star of the weighted adjacency matrix in a max-plus algebraic system. We focus on systems whose precedence constraints are represented by a directed acyclic graph and implement it on a Cell Broadband Engine™ (CBE) processor. Since the resulting matrix gives the longest travel times between two adjacent nodes, it is often utilized in scheduling problem solvers for a class of discrete event systems. This research, in particular, attempts to achieve a speedup by using two approaches: parallelization and SIMDization (Single Instruction, Multiple Data), both of which can be accomplished by a CBE processor. The former refers to a parallel computation using multiple cores, while the latter is a method whereby multiple elements are computed by a single instruction. Using the implementation on a Sony PlayStation 3™ equipped with a CBE processor, we found that the SIMDization is effective regardless of the system's size and the number of processor cores used. We also found that the scalability of using multiple cores is remarkable especially for systems with a large number of nodes. In a numerical experiment where the number of nodes is 2000, we achieved a speedup of 20 times compared with the method without the above techniques.

Investigations into the behaviour of Plasma surrounding Pulsars: DYMPHNA3D

NASA Astrophysics Data System (ADS)

Rochford, Ronan; Mc Donald, John; Shearer, Andy

2011-08-01

We report on a new 3D fully relativistic, modular, parallel and scalable Particle-In-Cell (PIC) code currently being developed at the Computational Astrophysics Laboratory in the National University of Ireland, Galway and its initial test applications to the plasma distribution in the vicinity of a rapidly rotating neutron star. We find that Plasma remains confined by trapping surfaces close to the star as opposed to propagating to a significant portion of the light-cylinder distance as predicted in this early work. We discuss planned future modifications and applications of the developed code.

Three-dimensional Finite Element Formulation and Scalable Domain Decomposition for High Fidelity Rotor Dynamic Analysis

NASA Technical Reports Server (NTRS)

Datta, Anubhav; Johnson, Wayne R.

2009-01-01

This paper has two objectives. The first objective is to formulate a 3-dimensional Finite Element Model for the dynamic analysis of helicopter rotor blades. The second objective is to implement and analyze a dual-primal iterative substructuring based Krylov solver, that is parallel and scalable, for the solution of the 3-D FEM analysis. The numerical and parallel scalability of the solver is studied using two prototype problems - one for ideal hover (symmetric) and one for a transient forward flight (non-symmetric) - both carried out on up to 48 processors. In both hover and forward flight conditions, a perfect linear speed-up is observed, for a given problem size, up to the point of substructure optimality. Substructure optimality and the linear parallel speed-up range are both shown to depend on the problem size as well as on the selection of the coarse problem. With a larger problem size, linear speed-up is restored up to the new substructure optimality. The solver also scales with problem size - even though this conclusion is premature given the small prototype grids considered in this study.

Using SpF to Achieve Petascale for Legacy Pseudospectral Applications

NASA Technical Reports Server (NTRS)

Clune, Thomas L.; Jiang, Weiyuan

2014-01-01

Pseudospectral (PS) methods possess a number of characteristics (e.g., efficiency, accuracy, natural boundary conditions) that are extremely desirable for dynamo models. Unfortunately, dynamo models based upon PS methods face a number of daunting challenges, which include exposing additional parallelism, leveraging hardware accelerators, exploiting hybrid parallelism, and improving the scalability of global memory transposes. Although these issues are a concern for most models, solutions for PS methods tend to require far more pervasive changes to underlying data and control structures. Further, improvements in performance in one model are difficult to transfer to other models, resulting in significant duplication of effort across the research community. We have developed an extensible software framework for pseudospectral methods called SpF that is intended to enable extreme scalability and optimal performance. Highlevel abstractions provided by SpF unburden applications of the responsibility of managing domain decomposition and load balance while reducing the changes in code required to adapt to new computing architectures. The key design concept in SpF is that each phase of the numerical calculation is partitioned into disjoint numerical kernels that can be performed entirely inprocessor. The granularity of domain decomposition provided by SpF is only constrained by the datalocality requirements of these kernels. SpF builds on top of optimized vendor libraries for common numerical operations such as transforms, matrix solvers, etc., but can also be configured to use open source alternatives for portability. SpF includes several alternative schemes for global data redistribution and is expected to serve as an ideal testbed for further research into optimal approaches for different network architectures. In this presentation, we will describe our experience in porting legacy pseudospectral models, MoSST and DYNAMO, to use SpF as well as present preliminary performance results provided by the improved scalability.

Computational Challenges of 3D Radiative Transfer in Atmospheric Models

NASA Astrophysics Data System (ADS)

Jakub, Fabian; Bernhard, Mayer

2017-04-01

The computation of radiative heating and cooling rates is one of the most expensive components in todays atmospheric models. The high computational cost stems not only from the laborious integration over a wide range of the electromagnetic spectrum but also from the fact that solving the integro-differential radiative transfer equation for monochromatic light is already rather involved. This lead to the advent of numerous approximations and parameterizations to reduce the cost of the solver. One of the most prominent one is the so called independent pixel approximations (IPA) where horizontal energy transfer is neglected whatsoever and radiation may only propagate in the vertical direction (1D). Recent studies implicate that the IPA introduces significant errors in high resolution simulations and affects the evolution and development of convective systems. However, using fully 3D solvers such as for example MonteCarlo methods is not even on state of the art supercomputers feasible. The parallelization of atmospheric models is often realized by a horizontal domain decomposition, and hence, horizontal transfer of energy necessitates communication. E.g. a cloud's shadow at a low zenith angle will cast a long shadow and potentially needs to communication through a multitude of processors. Especially light in the solar spectral range may travel long distances through the atmosphere. Concerning highly parallel simulations, it is vital that 3D radiative transfer solvers put a special emphasis on parallel scalability. We will present an introduction to intricacies computing 3D radiative heating and cooling rates as well as report on the parallel performance of the TenStream solver. The TenStream is a 3D radiative transfer solver using the PETSc framework to iteratively solve a set of partial differential equation. We investigate two matrix preconditioners, (a) geometric algebraic multigrid preconditioning(MG+GAMG) and (b) block Jacobi incomplete LU (ILU) factorization. The TenStream solver is tested for up to 4096 cores and shows a parallel scaling efficiency of 80-90% on various supercomputers.

Astronomy In The Cloud: Using Mapreduce For Image Coaddition

NASA Astrophysics Data System (ADS)

Wiley, Keith; Connolly, A.; Gardner, J.; Krughoff, S.; Balazinska, M.; Howe, B.; Kwon, Y.; Bu, Y.

2011-01-01

In the coming decade, astronomical surveys of the sky will generate tens of terabytes of images and detect hundreds of millions of sources every night. The study of these sources will involve computational challenges such as anomaly detection, classification, and moving object tracking. Since such studies require the highest quality data, methods such as image coaddition, i.e., registration, stacking, and mosaicing, will be critical to scientific investigation. With a requirement that these images be analyzed on a nightly basis to identify moving sources, e.g., asteroids, or transient objects, e.g., supernovae, these datastreams present many computational challenges. Given the quantity of data involved, the computational load of these problems can only be addressed by distributing the workload over a large number of nodes. However, the high data throughput demanded by these applications may present scalability challenges for certain storage architectures. One scalable data-processing method that has emerged in recent years is MapReduce, and in this paper we focus on its popular open-source implementation called Hadoop. In the Hadoop framework, the data is partitioned among storage attached directly to worker nodes, and the processing workload is scheduled in parallel on the nodes that contain the required input data. A further motivation for using Hadoop is that it allows us to exploit cloud computing resources, i.e., platforms where Hadoop is offered as a service. We report on our experience implementing a scalable image-processing pipeline for the SDSS imaging database using Hadoop. This multi-terabyte imaging dataset provides a good testbed for algorithm development since its scope and structure approximate future surveys. First, we describe MapReduce and how we adapted image coaddition to the MapReduce framework. Then we describe a number of optimizations to our basic approach and report experimental results compring their performance. This work is funded by the NSF and by NASA.

Astronomy in the Cloud: Using MapReduce for Image Co-Addition

NASA Astrophysics Data System (ADS)

Wiley, K.; Connolly, A.; Gardner, J.; Krughoff, S.; Balazinska, M.; Howe, B.; Kwon, Y.; Bu, Y.

2011-03-01

In the coming decade, astronomical surveys of the sky will generate tens of terabytes of images and detect hundreds of millions of sources every night. The study of these sources will involve computation challenges such as anomaly detection and classification and moving-object tracking. Since such studies benefit from the highest-quality data, methods such as image co-addition, i.e., astrometric registration followed by per-pixel summation, will be a critical preprocessing step prior to scientific investigation. With a requirement that these images be analyzed on a nightly basis to identify moving sources such as potentially hazardous asteroids or transient objects such as supernovae, these data streams present many computational challenges. Given the quantity of data involved, the computational load of these problems can only be addressed by distributing the workload over a large number of nodes. However, the high data throughput demanded by these applications may present scalability challenges for certain storage architectures. One scalable data-processing method that has emerged in recent years is MapReduce, and in this article we focus on its popular open-source implementation called Hadoop. In the Hadoop framework, the data are partitioned among storage attached directly to worker nodes, and the processing workload is scheduled in parallel on the nodes that contain the required input data. A further motivation for using Hadoop is that it allows us to exploit cloud computing resources: i.e., platforms where Hadoop is offered as a service. We report on our experience of implementing a scalable image-processing pipeline for the SDSS imaging database using Hadoop. This multiterabyte imaging data set provides a good testbed for algorithm development, since its scope and structure approximate future surveys. First, we describe MapReduce and how we adapted image co-addition to the MapReduce framework. Then we describe a number of optimizations to our basic approach and report experimental results comparing their performance.

Highly Parallel Alternating Directions Algorithm for Time Dependent Problems

NASA Astrophysics Data System (ADS)

Ganzha, M.; Georgiev, K.; Lirkov, I.; Margenov, S.; Paprzycki, M.

2011-11-01

In our work, we consider the time dependent Stokes equation on a finite time interval and on a uniform rectangular mesh, written in terms of velocity and pressure. For this problem, a parallel algorithm based on a novel direction splitting approach is developed. Here, the pressure equation is derived from a perturbed form of the continuity equation, in which the incompressibility constraint is penalized in a negative norm induced by the direction splitting. The scheme used in the algorithm is composed of two parts: (i) velocity prediction, and (ii) pressure correction. This is a Crank-Nicolson-type two-stage time integration scheme for two and three dimensional parabolic problems in which the second-order derivative, with respect to each space variable, is treated implicitly while the other variable is made explicit at each time sub-step. In order to achieve a good parallel performance the solution of the Poison problem for the pressure correction is replaced by solving a sequence of one-dimensional second order elliptic boundary value problems in each spatial direction. The parallel code is implemented using the standard MPI functions and tested on two modern parallel computer systems. The performed numerical tests demonstrate good level of parallel efficiency and scalability of the studied direction-splitting-based algorithm.

Scalable Visual Analytics of Massive Textual Datasets

DOE Office of Scientific and Technical Information (OSTI.GOV)

Krishnan, Manoj Kumar; Bohn, Shawn J.; Cowley, Wendy E.

2007-04-01

This paper describes the first scalable implementation of text processing engine used in Visual Analytics tools. These tools aid information analysts in interacting with and understanding large textual information content through visual interfaces. By developing parallel implementation of the text processing engine, we enabled visual analytics tools to exploit cluster architectures and handle massive dataset. The paper describes key elements of our parallelization approach and demonstrates virtually linear scaling when processing multi-gigabyte data sets such as Pubmed. This approach enables interactive analysis of large datasets beyond capabilities of existing state-of-the art visual analytics tools.

An Element-Based Concurrent Partitioner for Unstructured Finite Element Meshes

NASA Technical Reports Server (NTRS)

Ding, Hong Q.; Ferraro, Robert D.

1996-01-01

A concurrent partitioner for partitioning unstructured finite element meshes on distributed memory architectures is developed. The partitioner uses an element-based partitioning strategy. Its main advantage over the more conventional node-based partitioning strategy is its modular programming approach to the development of parallel applications. The partitioner first partitions element centroids using a recursive inertial bisection algorithm. Elements and nodes then migrate according to the partitioned centroids, using a data request communication template for unpredictable incoming messages. Our scalable implementation is contrasted to a non-scalable implementation which is a straightforward parallelization of a sequential partitioner.

Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures

DOE PAGES

Phipps, Eric T.; D'Elia, Marta; Edwards, Harold C.; ...

2017-04-18

In this study, quantifying simulation uncertainties is a critical component of rigorous predictive simulation. A key component of this is forward propagation of uncertainties in simulation input data to output quantities of interest. Typical approaches involve repeated sampling of the simulation over the uncertain input data, and can require numerous samples when accurately propagating uncertainties from large numbers of sources. Often simulation processes from sample to sample are similar and much of the data generated from each sample evaluation could be reused. We explore a new method for implementing sampling methods that simultaneously propagates groups of samples together in anmore » embedded fashion, which we call embedded ensemble propagation. We show how this approach takes advantage of properties of modern computer architectures to improve performance by enabling reuse between samples, reducing memory bandwidth requirements, improving memory access patterns, improving opportunities for fine-grained parallelization, and reducing communication costs. We describe a software technique for implementing embedded ensemble propagation based on the use of C++ templates and describe its integration with various scientific computing libraries within Trilinos. We demonstrate improved performance, portability and scalability for the approach applied to the simulation of partial differential equations on a variety of CPU, GPU, and accelerator architectures, including up to 131,072 cores on a Cray XK7 (Titan).« less

Modular Universal Scalable Ion-trap Quantum Computer

DTIC Science & Technology

2016-06-02

SECURITY CLASSIFICATION OF: The main goal of the original MUSIQC proposal was to construct and demonstrate a modular and universally- expandable ion...Distribution Unlimited UU UU UU UU 02-06-2016 1-Aug-2010 31-Jan-2016 Final Report: Modular Universal Scalable Ion-trap Quantum Computer The views...P.O. Box 12211 Research Triangle Park, NC 27709-2211 Ion trap quantum computation, scalable modular architectures REPORT DOCUMENTATION PAGE 11

Long-time dynamics through parallel trajectory splicing

DOE Office of Scientific and Technical Information (OSTI.GOV)

Perez, Danny; Cubuk, Ekin D.; Waterland, Amos

2015-11-24

Simulating the atomistic evolution of materials over long time scales is a longstanding challenge, especially for complex systems where the distribution of barrier heights is very heterogeneous. Such systems are difficult to investigate using conventional long-time scale techniques, and the fact that they tend to remain trapped in small regions of configuration space for extended periods of time strongly limits the physical insights gained from short simulations. We introduce a novel simulation technique, Parallel Trajectory Splicing (ParSplice), that aims at addressing this problem through the timewise parallelization of long trajectories. The computational efficiency of ParSplice stems from a speculation strategymore » whereby predictions of the future evolution of the system are leveraged to increase the amount of work that can be concurrently performed at any one time, hence improving the scalability of the method. ParSplice is also able to accurately account for, and potentially reuse, a substantial fraction of the computational work invested in the simulation. We validate the method on a simple Ag surface system and demonstrate substantial increases in efficiency compared to previous methods. As a result, we then demonstrate the power of ParSplice through the study of topology changes in Ag 42Cu 13 core–shell nanoparticles.« less

Developing a Hadoop-based Middleware for Handling Multi-dimensional NetCDF

NASA Astrophysics Data System (ADS)

Li, Z.; Yang, C. P.; Schnase, J. L.; Duffy, D.; Lee, T. J.

2014-12-01

Climate observations and model simulations are collecting and generating vast amounts of climate data, and these data are ever-increasing and being accumulated in a rapid speed. Effectively managing and analyzing these data are essential for climate change studies. Hadoop, a distributed storage and processing framework for large data sets, has attracted increasing attentions in dealing with the Big Data challenge. The maturity of Infrastructure as a Service (IaaS) of cloud computing further accelerates the adoption of Hadoop in solving Big Data problems. However, Hadoop is designed to process unstructured data such as texts, documents and web pages, and cannot effectively handle the scientific data format such as array-based NetCDF files and other binary data format. In this paper, we propose to build a Hadoop-based middleware for transparently handling big NetCDF data by 1) designing a distributed climate data storage mechanism based on POSIX-enabled parallel file system to enable parallel big data processing with MapReduce, as well as support data access by other systems; 2) modifying the Hadoop framework to transparently processing NetCDF data in parallel without sequencing or converting the data into other file formats, or loading them to HDFS; and 3) seamlessly integrating Hadoop, cloud computing and climate data in a highly scalable and fault-tolerance framework.

BigDebug: Debugging Primitives for Interactive Big Data Processing in Spark.

PubMed

Gulzar, Muhammad Ali; Interlandi, Matteo; Yoo, Seunghyun; Tetali, Sai Deep; Condie, Tyson; Millstein, Todd; Kim, Miryung

2016-05-01

Developers use cloud computing platforms to process a large quantity of data in parallel when developing big data analytics. Debugging the massive parallel computations that run in today's data-centers is time consuming and error-prone. To address this challenge, we design a set of interactive, real-time debugging primitives for big data processing in Apache Spark, the next generation data-intensive scalable cloud computing platform. This requires re-thinking the notion of step-through debugging in a traditional debugger such as gdb, because pausing the entire computation across distributed worker nodes causes significant delay and naively inspecting millions of records using a watchpoint is too time consuming for an end user. First, BIGDEBUG's simulated breakpoints and on-demand watchpoints allow users to selectively examine distributed, intermediate data on the cloud with little overhead. Second, a user can also pinpoint a crash-inducing record and selectively resume relevant sub-computations after a quick fix. Third, a user can determine the root causes of errors (or delays) at the level of individual records through a fine-grained data provenance capability. Our evaluation shows that BIGDEBUG scales to terabytes and its record-level tracing incurs less than 25% overhead on average. It determines crash culprits orders of magnitude more accurately and provides up to 100% time saving compared to the baseline replay debugger. The results show that BIGDEBUG supports debugging at interactive speeds with minimal performance impact.

Cloud Computing for Protein-Ligand Binding Site Comparison

PubMed Central

2013-01-01

The proteome-wide analysis of protein-ligand binding sites and their interactions with ligands is important in structure-based drug design and in understanding ligand cross reactivity and toxicity. The well-known and commonly used software, SMAP, has been designed for 3D ligand binding site comparison and similarity searching of a structural proteome. SMAP can also predict drug side effects and reassign existing drugs to new indications. However, the computing scale of SMAP is limited. We have developed a high availability, high performance system that expands the comparison scale of SMAP. This cloud computing service, called Cloud-PLBS, combines the SMAP and Hadoop frameworks and is deployed on a virtual cloud computing platform. To handle the vast amount of experimental data on protein-ligand binding site pairs, Cloud-PLBS exploits the MapReduce paradigm as a management and parallelizing tool. Cloud-PLBS provides a web portal and scalability through which biologists can address a wide range of computer-intensive questions in biology and drug discovery. PMID:23762824

Cloud computing for protein-ligand binding site comparison.

PubMed

Hung, Che-Lun; Hua, Guan-Jie

2013-01-01

The proteome-wide analysis of protein-ligand binding sites and their interactions with ligands is important in structure-based drug design and in understanding ligand cross reactivity and toxicity. The well-known and commonly used software, SMAP, has been designed for 3D ligand binding site comparison and similarity searching of a structural proteome. SMAP can also predict drug side effects and reassign existing drugs to new indications. However, the computing scale of SMAP is limited. We have developed a high availability, high performance system that expands the comparison scale of SMAP. This cloud computing service, called Cloud-PLBS, combines the SMAP and Hadoop frameworks and is deployed on a virtual cloud computing platform. To handle the vast amount of experimental data on protein-ligand binding site pairs, Cloud-PLBS exploits the MapReduce paradigm as a management and parallelizing tool. Cloud-PLBS provides a web portal and scalability through which biologists can address a wide range of computer-intensive questions in biology and drug discovery.

Binary Interval Search: a scalable algorithm for counting interval intersections.

PubMed

Layer, Ryan M; Skadron, Kevin; Robins, Gabriel; Hall, Ira M; Quinlan, Aaron R

2013-01-01

The comparison of diverse genomic datasets is fundamental to understand genome biology. Researchers must explore many large datasets of genome intervals (e.g. genes, sequence alignments) to place their experimental results in a broader context and to make new discoveries. Relationships between genomic datasets are typically measured by identifying intervals that intersect, that is, they overlap and thus share a common genome interval. Given the continued advances in DNA sequencing technologies, efficient methods for measuring statistically significant relationships between many sets of genomic features are crucial for future discovery. We introduce the Binary Interval Search (BITS) algorithm, a novel and scalable approach to interval set intersection. We demonstrate that BITS outperforms existing methods at counting interval intersections. Moreover, we show that BITS is intrinsically suited to parallel computing architectures, such as graphics processing units by illustrating its utility for efficient Monte Carlo simulations measuring the significance of relationships between sets of genomic intervals. https://github.com/arq5x/bits.

JuxtaView - A tool for interactive visualization of large imagery on scalable tiled displays

USGS Publications Warehouse

Krishnaprasad, N.K.; Vishwanath, V.; Venkataraman, S.; Rao, A.G.; Renambot, L.; Leigh, J.; Johnson, A.E.; Davis, B.

2004-01-01

JuxtaView is a cluster-based application for viewing ultra-high-resolution images on scalable tiled displays. We present in JuxtaView, a new parallel computing and distributed memory approach for out-of-core montage visualization, using LambdaRAM, a software-based network-level cache system. The ultimate goal of JuxtaView is to enable a user to interactively roam through potentially terabytes of distributed, spatially referenced image data such as those from electron microscopes, satellites and aerial photographs. In working towards this goal, we describe our first prototype implemented over a local area network, where the image is distributed using LambdaRAM, on the memory of all nodes of a PC cluster driving a tiled display wall. Aggressive pre-fetching schemes employed by LambdaRAM help to reduce latency involved in remote memory access. We compare LambdaRAM with a more traditional memory-mapped file approach for out-of-core visualization. ?? 2004 IEEE.

MarFS, a Near-POSIX Interface to Cloud Objects

DOE Office of Scientific and Technical Information (OSTI.GOV)

Inman, Jeffrey Thornton; Vining, William Flynn; Ransom, Garrett Wilson

The engineering forces driving development of “cloud” storage have produced resilient, cost-effective storage systems that can scale to 100s of petabytes, with good parallel access and bandwidth. These features would make a good match for the vast storage needs of High-Performance Computing datacenters, but cloud storage gains some of its capability from its use of HTTP-style Representational State Transfer (REST) semantics, whereas most large datacenters have legacy applications that rely on POSIX file-system semantics. MarFS is an open-source project at Los Alamos National Laboratory that allows us to present cloud-style object-storage as a scalable near-POSIX file system. We have alsomore » developed a new storage architecture to improve bandwidth and scalability beyond what’s available in commodity object stores, while retaining their resilience and economy. Additionally, we present a scheme for scaling the POSIX interface to allow billions of files in a single directory and trillions of files in total.« less

MarFS, a Near-POSIX Interface to Cloud Objects

DOE PAGES

Inman, Jeffrey Thornton; Vining, William Flynn; Ransom, Garrett Wilson; ...

2017-01-01

The engineering forces driving development of “cloud” storage have produced resilient, cost-effective storage systems that can scale to 100s of petabytes, with good parallel access and bandwidth. These features would make a good match for the vast storage needs of High-Performance Computing datacenters, but cloud storage gains some of its capability from its use of HTTP-style Representational State Transfer (REST) semantics, whereas most large datacenters have legacy applications that rely on POSIX file-system semantics. MarFS is an open-source project at Los Alamos National Laboratory that allows us to present cloud-style object-storage as a scalable near-POSIX file system. We have alsomore » developed a new storage architecture to improve bandwidth and scalability beyond what’s available in commodity object stores, while retaining their resilience and economy. Additionally, we present a scheme for scaling the POSIX interface to allow billions of files in a single directory and trillions of files in total.« less

Highly parallel implementation of non-adiabatic Ehrenfest molecular dynamics

NASA Astrophysics Data System (ADS)

Kanai, Yosuke; Schleife, Andre; Draeger, Erik; Anisimov, Victor; Correa, Alfredo

2014-03-01

While the adiabatic Born-Oppenheimer approximation tremendously lowers computational effort, many questions in modern physics, chemistry, and materials science require an explicit description of coupled non-adiabatic electron-ion dynamics. Electronic stopping, i.e. the energy transfer of a fast projectile atom to the electronic system of the target material, is a notorious example. We recently implemented real-time time-dependent density functional theory based on the plane-wave pseudopotential formalism in the Qbox/qb@ll codes. We demonstrate that explicit integration using a fourth-order Runge-Kutta scheme is very suitable for modern highly parallelized supercomputers. Applying the new implementation to systems with hundreds of atoms and thousands of electrons, we achieved excellent performance and scalability on a large number of nodes both on the BlueGene based ``Sequoia'' system at LLNL as well as the Cray architecture of ``Blue Waters'' at NCSA. As an example, we discuss our work on computing the electronic stopping power of aluminum and gold for hydrogen projectiles, showing an excellent agreement with experiment. These first-principles calculations allow us to gain important insight into the the fundamental physics of electronic stopping.

A multi-platform evaluation of the randomized CX low-rank matrix factorization in Spark

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gittens, Alex; Kottalam, Jey; Yang, Jiyan

We investigate the performance and scalability of the randomized CX low-rank matrix factorization and demonstrate its applicability through the analysis of a 1TB mass spectrometry imaging (MSI) dataset, using Apache Spark on an Amazon EC2 cluster, a Cray XC40 system, and an experimental Cray cluster. We implemented this factorization both as a parallelized C implementation with hand-tuned optimizations and in Scala using the Apache Spark high-level cluster computing framework. We obtained consistent performance across the three platforms: using Spark we were able to process the 1TB size dataset in under 30 minutes with 960 cores on all systems, with themore » fastest times obtained on the experimental Cray cluster. In comparison, the C implementation was 21X faster on the Amazon EC2 system, due to careful cache optimizations, bandwidth-friendly access of matrices and vector computation using SIMD units. We report these results and their implications on the hardware and software issues arising in supporting data-centric workloads in parallel and distributed environments.« less

Avoiding and tolerating latency in large-scale next-generation shared-memory multiprocessors

NASA Technical Reports Server (NTRS)

Probst, David K.

1993-01-01

A scalable solution to the memory-latency problem is necessary to prevent the large latencies of synchronization and memory operations inherent in large-scale shared-memory multiprocessors from reducing high performance. We distinguish latency avoidance and latency tolerance. Latency is avoided when data is brought to nearby locales for future reference. Latency is tolerated when references are overlapped with other computation. Latency-avoiding locales include: processor registers, data caches used temporally, and nearby memory modules. Tolerating communication latency requires parallelism, allowing the overlap of communication and computation. Latency-tolerating techniques include: vector pipelining, data caches used spatially, prefetching in various forms, and multithreading in various forms. Relaxing the consistency model permits increased use of avoidance and tolerance techniques. Each model is a mapping from the program text to sets of partial orders on program operations; it is a convention about which temporal precedences among program operations are necessary. Information about temporal locality and parallelism constrains the use of avoidance and tolerance techniques. Suitable architectural primitives and compiler technology are required to exploit the increased freedom to reorder and overlap operations in relaxed models.

Scalable Unix commands for parallel processors : a high-performance implementation.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ong, E.; Lusk, E.; Gropp, W.

2001-06-22

We describe a family of MPI applications we call the Parallel Unix Commands. These commands are natural parallel versions of common Unix user commands such as ls, ps, and find, together with a few similar commands particular to the parallel environment. We describe the design and implementation of these programs and present some performance results on a 256-node Linux cluster. The Parallel Unix Commands are open source and freely available.

Scalable problems and memory bounded speedup

NASA Technical Reports Server (NTRS)

Sun, Xian-He; Ni, Lionel M.

1992-01-01

In this paper three models of parallel speedup are studied. They are fixed-size speedup, fixed-time speedup and memory-bounded speedup. The latter two consider the relationship between speedup and problem scalability. Two sets of speedup formulations are derived for these three models. One set considers uneven workload allocation and communication overhead and gives more accurate estimation. Another set considers a simplified case and provides a clear picture on the impact of the sequential portion of an application on the possible performance gain from parallel processing. The simplified fixed-size speedup is Amdahl's law. The simplified fixed-time speedup is Gustafson's scaled speedup. The simplified memory-bounded speedup contains both Amdahl's law and Gustafson's scaled speedup as special cases. This study leads to a better understanding of parallel processing.

A scalable neuroinformatics data flow for electrophysiological signals using MapReduce.

PubMed

Jayapandian, Catherine; Wei, Annan; Ramesh, Priya; Zonjy, Bilal; Lhatoo, Samden D; Loparo, Kenneth; Zhang, Guo-Qiang; Sahoo, Satya S

2015-01-01

Data-driven neuroscience research is providing new insights in progression of neurological disorders and supporting the development of improved treatment approaches. However, the volume, velocity, and variety of neuroscience data generated from sophisticated recording instruments and acquisition methods have exacerbated the limited scalability of existing neuroinformatics tools. This makes it difficult for neuroscience researchers to effectively leverage the growing multi-modal neuroscience data to advance research in serious neurological disorders, such as epilepsy. We describe the development of the Cloudwave data flow that uses new data partitioning techniques to store and analyze electrophysiological signal in distributed computing infrastructure. The Cloudwave data flow uses MapReduce parallel programming algorithm to implement an integrated signal data processing pipeline that scales with large volume of data generated at high velocity. Using an epilepsy domain ontology together with an epilepsy focused extensible data representation format called Cloudwave Signal Format (CSF), the data flow addresses the challenge of data heterogeneity and is interoperable with existing neuroinformatics data representation formats, such as HDF5. The scalability of the Cloudwave data flow is evaluated using a 30-node cluster installed with the open source Hadoop software stack. The results demonstrate that the Cloudwave data flow can process increasing volume of signal data by leveraging Hadoop Data Nodes to reduce the total data processing time. The Cloudwave data flow is a template for developing highly scalable neuroscience data processing pipelines using MapReduce algorithms to support a variety of user applications.

A scalable neuroinformatics data flow for electrophysiological signals using MapReduce

PubMed Central

Jayapandian, Catherine; Wei, Annan; Ramesh, Priya; Zonjy, Bilal; Lhatoo, Samden D.; Loparo, Kenneth; Zhang, Guo-Qiang; Sahoo, Satya S.

2015-01-01

Data-driven neuroscience research is providing new insights in progression of neurological disorders and supporting the development of improved treatment approaches. However, the volume, velocity, and variety of neuroscience data generated from sophisticated recording instruments and acquisition methods have exacerbated the limited scalability of existing neuroinformatics tools. This makes it difficult for neuroscience researchers to effectively leverage the growing multi-modal neuroscience data to advance research in serious neurological disorders, such as epilepsy. We describe the development of the Cloudwave data flow that uses new data partitioning techniques to store and analyze electrophysiological signal in distributed computing infrastructure. The Cloudwave data flow uses MapReduce parallel programming algorithm to implement an integrated signal data processing pipeline that scales with large volume of data generated at high velocity. Using an epilepsy domain ontology together with an epilepsy focused extensible data representation format called Cloudwave Signal Format (CSF), the data flow addresses the challenge of data heterogeneity and is interoperable with existing neuroinformatics data representation formats, such as HDF5. The scalability of the Cloudwave data flow is evaluated using a 30-node cluster installed with the open source Hadoop software stack. The results demonstrate that the Cloudwave data flow can process increasing volume of signal data by leveraging Hadoop Data Nodes to reduce the total data processing time. The Cloudwave data flow is a template for developing highly scalable neuroscience data processing pipelines using MapReduce algorithms to support a variety of user applications. PMID:25852536

A Scalable, Parallel Approach for Multi-Point, High-Fidelity Aerostructural Optimization of Aircraft Configurations

NASA Astrophysics Data System (ADS)

Kenway, Gaetan K. W.

This thesis presents new tools and techniques developed to address the challenging problem of high-fidelity aerostructural optimization with respect to large numbers of design variables. A new mesh-movement scheme is developed that is both computationally efficient and sufficiently robust to accommodate large geometric design changes and aerostructural deformations. A fully coupled Newton-Krylov method is presented that accelerates the convergence of aerostructural systems and provides a 20% performance improvement over the traditional nonlinear block Gauss-Seidel approach and can handle more exible structures. A coupled adjoint method is used that efficiently computes derivatives for a gradient-based optimization algorithm. The implementation uses only machine accurate derivative techniques and is verified to yield fully consistent derivatives by comparing against the complex step method. The fully-coupled large-scale coupled adjoint solution method is shown to have 30% better performance than the segregated approach. The parallel scalability of the coupled adjoint technique is demonstrated on an Euler Computational Fluid Dynamics (CFD) model with more than 80 million state variables coupled to a detailed structural finite-element model of the wing with more than 1 million degrees of freedom. Multi-point high-fidelity aerostructural optimizations of a long-range wide-body, transonic transport aircraft configuration are performed using the developed techniques. The aerostructural analysis employs Euler CFD with a 2 million cell mesh and a structural finite element model with 300 000 DOF. Two design optimization problems are solved: one where takeoff gross weight is minimized, and another where fuel burn is minimized. Each optimization uses a multi-point formulation with 5 cruise conditions and 2 maneuver conditions. The optimization problems have 476 design variables are optimal results are obtained within 36 hours of wall time using 435 processors. The TOGW minimization results in a 4.2% reduction in TOGW with a 6.6% fuel burn reduction, while the fuel burn optimization resulted in a 11.2% fuel burn reduction with no change to the takeoff gross weight.

Implementation of a flexible and scalable particle-in-cell method for massively parallel computations in the mantle convection code ASPECT

NASA Astrophysics Data System (ADS)

Gassmöller, Rene; Bangerth, Wolfgang

2016-04-01

Particle-in-cell methods have a long history and many applications in geodynamic modelling of mantle convection, lithospheric deformation and crustal dynamics. They are primarily used to track material information, the strain a material has undergone, the pressure-temperature history a certain material region has experienced, or the amount of volatiles or partial melt present in a region. However, their efficient parallel implementation - in particular combined with adaptive finite-element meshes - is complicated due to the complex communication patterns and frequent reassignment of particles to cells. Consequently, many current scientific software packages accomplish this efficient implementation by specifically designing particle methods for a single purpose, like the advection of scalar material properties that do not evolve over time (e.g., for chemical heterogeneities). Design choices for particle integration, data storage, and parallel communication are then optimized for this single purpose, making the code relatively rigid to changing requirements. Here, we present the implementation of a flexible, scalable and efficient particle-in-cell method for massively parallel finite-element codes with adaptively changing meshes. Using a modular plugin structure, we allow maximum flexibility of the generation of particles, the carried tracer properties, the advection and output algorithms, and the projection of properties to the finite-element mesh. We present scaling tests ranging up to tens of thousands of cores and tens of billions of particles. Additionally, we discuss efficient load-balancing strategies for particles in adaptive meshes with their strengths and weaknesses, local particle-transfer between parallel subdomains utilizing existing communication patterns from the finite element mesh, and the use of established parallel output algorithms like the HDF5 library. Finally, we show some relevant particle application cases, compare our implementation to a modern advection-field approach, and demonstrate under which conditions which method is more efficient. We implemented the presented methods in ASPECT (aspect.dealii.org), a freely available open-source community code for geodynamic simulations. The structure of the particle code is highly modular, and segregated from the PDE solver, and can thus be easily transferred to other programs, or adapted for various application cases.

geoKepler Workflow Module for Computationally Scalable and Reproducible Geoprocessing and Modeling

NASA Astrophysics Data System (ADS)

Cowart, C.; Block, J.; Crawl, D.; Graham, J.; Gupta, A.; Nguyen, M.; de Callafon, R.; Smarr, L.; Altintas, I.

2015-12-01

The NSF-funded WIFIRE project has developed an open-source, online geospatial workflow platform for unifying geoprocessing tools and models for for fire and other geospatially dependent modeling applications. It is a product of WIFIRE's objective to build an end-to-end cyberinfrastructure for real-time and data-driven simulation, prediction and visualization of wildfire behavior. geoKepler includes a set of reusable GIS components, or actors, for the Kepler Scientific Workflow System (https://kepler-project.org). Actors exist for reading and writing GIS data in formats such as Shapefile, GeoJSON, KML, and using OGC web services such as WFS. The actors also allow for calling geoprocessing tools in other packages such as GDAL and GRASS. Kepler integrates functions from multiple platforms and file formats into one framework, thus enabling optimal GIS interoperability, model coupling, and scalability. Products of the GIS actors can be fed directly to models such as FARSITE and WRF. Kepler's ability to schedule and scale processes using Hadoop and Spark also makes geoprocessing ultimately extensible and computationally scalable. The reusable workflows in geoKepler can be made to run automatically when alerted by real-time environmental conditions. Here, we show breakthroughs in the speed of creating complex data for hazard assessments with this platform. We also demonstrate geoKepler workflows that use Data Assimilation to ingest real-time weather data into wildfire simulations, and for data mining techniques to gain insight into environmental conditions affecting fire behavior. Existing machine learning tools and libraries such as R and MLlib are being leveraged for this purpose in Kepler, as well as Kepler's Distributed Data Parallel (DDP) capability to provide a framework for scalable processing. geoKepler workflows can be executed via an iPython notebook as a part of a Jupyter hub at UC San Diego for sharing and reporting of the scientific analysis and results from various runs of geoKepler workflows. The communication between iPython and Kepler workflow executions is established through an iPython magic function for Kepler that we have implemented. In summary, geoKepler is an ecosystem that makes geospatial processing and analysis of any kind programmable, reusable, scalable and sharable.

Implementation of the DPM Monte Carlo code on a parallel architecture for treatment planning applications.

PubMed

Tyagi, Neelam; Bose, Abhijit; Chetty, Indrin J

2004-09-01

We have parallelized the Dose Planning Method (DPM), a Monte Carlo code optimized for radiotherapy class problems, on distributed-memory processor architectures using the Message Passing Interface (MPI). Parallelization has been investigated on a variety of parallel computing architectures at the University of Michigan-Center for Advanced Computing, with respect to efficiency and speedup as a function of the number of processors. We have integrated the parallel pseudo random number generator from the Scalable Parallel Pseudo-Random Number Generator (SPRNG) library to run with the parallel DPM. The Intel cluster consisting of 800 MHz Intel Pentium III processor shows an almost linear speedup up to 32 processors for simulating 1 x 10(8) or more particles. The speedup results are nearly linear on an Athlon cluster (up to 24 processors based on availability) which consists of 1.8 GHz+ Advanced Micro Devices (AMD) Athlon processors on increasing the problem size up to 8 x 10(8) histories. For a smaller number of histories (1 x 10(8)) the reduction of efficiency with the Athlon cluster (down to 83.9% with 24 processors) occurs because the processing time required to simulate 1 x 10(8) histories is less than the time associated with interprocessor communication. A similar trend was seen with the Opteron Cluster (consisting of 1400 MHz, 64-bit AMD Opteron processors) on increasing the problem size. Because of the 64-bit architecture Opteron processors are capable of storing and processing instructions at a faster rate and hence are faster as compared to the 32-bit Athlon processors. We have validated our implementation with an in-phantom dose calculation study using a parallel pencil monoenergetic electron beam of 20 MeV energy. The phantom consists of layers of water, lung, bone, aluminum, and titanium. The agreement in the central axis depth dose curves and profiles at different depths shows that the serial and parallel codes are equivalent in accuracy.

Near-realtime simulations of biolelectric activity in small mammalian hearts using graphical processing units

PubMed Central

Vigmond, Edward J.; Boyle, Patrick M.; Leon, L. Joshua; Plank, Gernot

2014-01-01

Simulations of cardiac bioelectric phenomena remain a significant challenge despite continual advancements in computational machinery. Spanning large temporal and spatial ranges demands millions of nodes to accurately depict geometry, and a comparable number of timesteps to capture dynamics. This study explores a new hardware computing paradigm, the graphics processing unit (GPU), to accelerate cardiac models, and analyzes results in the context of simulating a small mammalian heart in real time. The ODEs associated with membrane ionic flow were computed on traditional CPU and compared to GPU performance, for one to four parallel processing units. The scalability of solving the PDE responsible for tissue coupling was examined on a cluster using up to 128 cores. Results indicate that the GPU implementation was between 9 and 17 times faster than the CPU implementation and scaled similarly. Solving the PDE was still 160 times slower than real time. PMID:19964295

Launching genomics into the cloud: deployment of Mercury, a next generation sequence analysis pipeline.

PubMed

Reid, Jeffrey G; Carroll, Andrew; Veeraraghavan, Narayanan; Dahdouli, Mahmoud; Sundquist, Andreas; English, Adam; Bainbridge, Matthew; White, Simon; Salerno, William; Buhay, Christian; Yu, Fuli; Muzny, Donna; Daly, Richard; Duyk, Geoff; Gibbs, Richard A; Boerwinkle, Eric

2014-01-29

Massively parallel DNA sequencing generates staggering amounts of data. Decreasing cost, increasing throughput, and improved annotation have expanded the diversity of genomics applications in research and clinical practice. This expanding scale creates analytical challenges: accommodating peak compute demand, coordinating secure access for multiple analysts, and sharing validated tools and results. To address these challenges, we have developed the Mercury analysis pipeline and deployed it in local hardware and the Amazon Web Services cloud via the DNAnexus platform. Mercury is an automated, flexible, and extensible analysis workflow that provides accurate and reproducible genomic results at scales ranging from individuals to large cohorts. By taking advantage of cloud computing and with Mercury implemented on the DNAnexus platform, we have demonstrated a powerful combination of a robust and fully validated software pipeline and a scalable computational resource that, to date, we have applied to more than 10,000 whole genome and whole exome samples.

2HOT: An Improved Parallel Hashed Oct-Tree N-Body Algorithm for Cosmological Simulation

DOE PAGES

Warren, Michael S.

2014-01-01

We report on improvements made over the past two decades to our adaptive treecode N-body method (HOT). A mathematical and computational approach to the cosmological N-body problem is described, with performance and scalability measured up to 256k (2 18 ) processors. We present error analysis and scientific application results from a series of more than ten 69 billion (4096 3 ) particle cosmological simulations, accounting for 4×10 20 floating point operations. These results include the first simulations using the new constraints on the standard model of cosmology from the Planck satellite. Our simulations set a new standard for accuracy andmore » scientific throughput, while meeting or exceeding the computational efficiency of the latest generation of hybrid TreePM N-body methods.« less

FPGA-based coprocessor for matrix algorithms implementation

NASA Astrophysics Data System (ADS)

Amira, Abbes; Bensaali, Faycal

2003-03-01

Matrix algorithms are important in many types of applications including image and signal processing. These areas require enormous computing power. A close examination of the algorithms used in these, and related, applications reveals that many of the fundamental actions involve matrix operations such as matrix multiplication which is of O (N3) on a sequential computer and O (N3/p) on a parallel system with p processors complexity. This paper presents an investigation into the design and implementation of different matrix algorithms such as matrix operations, matrix transforms and matrix decompositions using an FPGA based environment. Solutions for the problem of processing large matrices have been proposed. The proposed system architectures are scalable, modular and require less area and time complexity with reduced latency when compared with existing structures.

Efficient computation of k-Nearest Neighbour Graphs for large high-dimensional data sets on GPU clusters.

PubMed

Dashti, Ali; Komarov, Ivan; D'Souza, Roshan M

2013-01-01

This paper presents an implementation of the brute-force exact k-Nearest Neighbor Graph (k-NNG) construction for ultra-large high-dimensional data cloud. The proposed method uses Graphics Processing Units (GPUs) and is scalable with multi-levels of parallelism (between nodes of a cluster, between different GPUs on a single node, and within a GPU). The method is applicable to homogeneous computing clusters with a varying number of nodes and GPUs per node. We achieve a 6-fold speedup in data processing as compared with an optimized method running on a cluster of CPUs and bring a hitherto impossible [Formula: see text]-NNG generation for a dataset of twenty million images with 15 k dimensionality into the realm of practical possibility.

Opportunities and choice in a new vector era

NASA Astrophysics Data System (ADS)

Nowak, A.

2014-06-01

This work discusses the significant changes in computing landscape related to the progression of Moore's Law, and the implications on scientific computing. Particular attention is devoted to the High Energy Physics domain (HEP), which has always made good use of threading, but levels of parallelism closer to the hardware were often left underutilized. Findings of the CERN openlab Platform Competence Center are reported in the context of expanding "performance dimensions", and especially the resurgence of vectors. These suggest that data oriented designs are feasible in HEP and have considerable potential for performance improvements on multiple levels, but will rarely trump algorithmic enhancements. Finally, an analysis of upcoming hardware and software technologies identifies heterogeneity as a major challenge for software, which will require more emphasis on scalable, efficient design.

Argonne simulation framework for intelligent transportation systems

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ewing, T.; Doss, E.; Hanebutte, U.

1996-04-01

A simulation framework has been developed which defines a high-level architecture for a large-scale, comprehensive, scalable simulation of an Intelligent Transportation System (ITS). The simulator is designed to run on parallel computers and distributed (networked) computer systems; however, a version for a stand alone workstation is also available. The ITS simulator includes an Expert Driver Model (EDM) of instrumented ``smart`` vehicles with in-vehicle navigation units. The EDM is capable of performing optimal route planning and communicating with Traffic Management Centers (TMC). A dynamic road map data base is sued for optimum route planning, where the data is updated periodically tomore » reflect any changes in road or weather conditions. The TMC has probe vehicle tracking capabilities (display position and attributes of instrumented vehicles), and can provide 2-way interaction with traffic to provide advisories and link times. Both the in-vehicle navigation module and the TMC feature detailed graphical user interfaces that includes human-factors studies to support safety and operational research. Realistic modeling of variations of the posted driving speed are based on human factor studies that take into consideration weather, road conditions, driver`s personality and behavior and vehicle type. The simulator has been developed on a distributed system of networked UNIX computers, but is designed to run on ANL`s IBM SP-X parallel computer system for large scale problems. A novel feature of the developed simulator is that vehicles will be represented by autonomous computer processes, each with a behavior model which performs independent route selection and reacts to external traffic events much like real vehicles. Vehicle processes interact with each other and with ITS components by exchanging messages. With this approach, one will be able to take advantage of emerging massively parallel processor (MPP) systems.« less

The Automated Instrumentation and Monitoring System (AIMS) reference manual

NASA Technical Reports Server (NTRS)

Yan, Jerry; Hontalas, Philip; Listgarten, Sherry

1993-01-01

Whether a researcher is designing the 'next parallel programming paradigm,' another 'scalable multiprocessor' or investigating resource allocation algorithms for multiprocessors, a facility that enables parallel program execution to be captured and displayed is invaluable. Careful analysis of execution traces can help computer designers and software architects to uncover system behavior and to take advantage of specific application characteristics and hardware features. A software tool kit that facilitates performance evaluation of parallel applications on multiprocessors is described. The Automated Instrumentation and Monitoring System (AIMS) has four major software components: a source code instrumentor which automatically inserts active event recorders into the program's source code before compilation; a run time performance-monitoring library, which collects performance data; a trace file animation and analysis tool kit which reconstructs program execution from the trace file; and a trace post-processor which compensate for data collection overhead. Besides being used as prototype for developing new techniques for instrumenting, monitoring, and visualizing parallel program execution, AIMS is also being incorporated into the run-time environments of various hardware test beds to evaluate their impact on user productivity. Currently, AIMS instrumentors accept FORTRAN and C parallel programs written for Intel's NX operating system on the iPSC family of multi computers. A run-time performance-monitoring library for the iPSC/860 is included in this release. We plan to release monitors for other platforms (such as PVM and TMC's CM-5) in the near future. Performance data collected can be graphically displayed on workstations (e.g. Sun Sparc and SGI) supporting X-Windows (in particular, Xl IR5, Motif 1.1.3).

Parallel reduced-instruction-set-computer architecture for real-time symbolic pattern matching

NASA Astrophysics Data System (ADS)

Parson, Dale E.

1991-03-01

This report discusses ongoing work on a parallel reduced-instruction- set-computer (RISC) architecture for automatic production matching. The PRIOPS compiler takes advantage of the memoryless character of automatic processing by translating a program's collection of automatic production tests into an equivalent combinational circuit-a digital circuit without memory, whose outputs are immediate functions of its inputs. The circuit provides a highly parallel, fine-grain model of automatic matching. The compiler then maps the combinational circuit onto RISC hardware. The heart of the processor is an array of comparators capable of testing production conditions in parallel, Each comparator attaches to private memory that contains virtual circuit nodes-records of the current state of nodes and busses in the combinational circuit. All comparator memories hold identical information, allowing simultaneous update for a single changing circuit node and simultaneous retrieval of different circuit nodes by different comparators. Along with the comparator-based logic unit is a sequencer that determines the current combination of production-derived comparisons to try, based on the combined success and failure of previous combinations of comparisons. The memoryless nature of automatic matching allows the compiler to designate invariant memory addresses for virtual circuit nodes, and to generate the most effective sequences of comparison test combinations. The result is maximal utilization of parallel hardware, indicating speed increases and scalability beyond that found for course-grain, multiprocessor approaches to concurrent Rete matching. Future work will consider application of this RISC architecture to the standard (controlled) Rete algorithm, where search through memory dominates portions of matching.

Towards Scalable Deep Learning via I/O Analysis and Optimization

DOE Office of Scientific and Technical Information (OSTI.GOV)

Pumma, Sarunya; Si, Min; Feng, Wu-Chun

Deep learning systems have been growing in prominence as a way to automatically characterize objects, trends, and anomalies. Given the importance of deep learning systems, researchers have been investigating techniques to optimize such systems. An area of particular interest has been using large supercomputing systems to quickly generate effective deep learning networks: a phase often referred to as “training” of the deep learning neural network. As we scale existing deep learning frameworks—such as Caffe—on these large supercomputing systems, we notice that the parallelism can help improve the computation tremendously, leaving data I/O as the major bottleneck limiting the overall systemmore » scalability. In this paper, we first present a detailed analysis of the performance bottlenecks of Caffe on large supercomputing systems. Our analysis shows that the I/O subsystem of Caffe—LMDB—relies on memory-mapped I/O to access its database, which can be highly inefficient on large-scale systems because of its interaction with the process scheduling system and the network-based parallel filesystem. Based on this analysis, we then present LMDBIO, our optimized I/O plugin for Caffe that takes into account the data access pattern of Caffe in order to vastly improve I/O performance. Our experimental results show that LMDBIO can improve the overall execution time of Caffe by nearly 20-fold in some cases.« less

Efficient parallel and out of core algorithms for constructing large bi-directed de Bruijn graphs.

PubMed

Kundeti, Vamsi K; Rajasekaran, Sanguthevar; Dinh, Hieu; Vaughn, Matthew; Thapar, Vishal

2010-11-15

Assembling genomic sequences from a set of overlapping reads is one of the most fundamental problems in computational biology. Algorithms addressing the assembly problem fall into two broad categories - based on the data structures which they employ. The first class uses an overlap/string graph and the second type uses a de Bruijn graph. However with the recent advances in short read sequencing technology, de Bruijn graph based algorithms seem to play a vital role in practice. Efficient algorithms for building these massive de Bruijn graphs are very essential in large sequencing projects based on short reads. In an earlier work, an O(n/p) time parallel algorithm has been given for this problem. Here n is the size of the input and p is the number of processors. This algorithm enumerates all possible bi-directed edges which can overlap with a node and ends up generating Θ(nΣ) messages (Σ being the size of the alphabet). In this paper we present a Θ(n/p) time parallel algorithm with a communication complexity that is equal to that of parallel sorting and is not sensitive to Σ. The generality of our algorithm makes it very easy to extend it even to the out-of-core model and in this case it has an optimal I/O complexity of Θ(nlog(n/B)Blog(M/B)) (M being the main memory size and B being the size of the disk block). We demonstrate the scalability of our parallel algorithm on a SGI/Altix computer. A comparison of our algorithm with the previous approaches reveals that our algorithm is faster--both asymptotically and practically. We demonstrate the scalability of our sequential out-of-core algorithm by comparing it with the algorithm used by VELVET to build the bi-directed de Bruijn graph. Our experiments reveal that our algorithm can build the graph with a constant amount of memory, which clearly outperforms VELVET. We also provide efficient algorithms for the bi-directed chain compaction problem. The bi-directed de Bruijn graph is a fundamental data structure for any sequence assembly program based on Eulerian approach. Our algorithms for constructing Bi-directed de Bruijn graphs are efficient in parallel and out of core settings. These algorithms can be used in building large scale bi-directed de Bruijn graphs. Furthermore, our algorithms do not employ any all-to-all communications in a parallel setting and perform better than the prior algorithms. Finally our out-of-core algorithm is extremely memory efficient and can replace the existing graph construction algorithm in VELVET.

Genetic algorithm based task reordering to improve the performance of batch scheduled massively parallel scientific applications

DOE PAGES

Sankaran, Ramanan; Angel, Jordan; Brown, W. Michael

2015-04-08

The growth in size of networked high performance computers along with novel accelerator-based node architectures has further emphasized the importance of communication efficiency in high performance computing. The world's largest high performance computers are usually operated as shared user facilities due to the costs of acquisition and operation. Applications are scheduled for execution in a shared environment and are placed on nodes that are not necessarily contiguous on the interconnect. Furthermore, the placement of tasks on the nodes allocated by the scheduler is sub-optimal, leading to performance loss and variability. Here, we investigate the impact of task placement on themore » performance of two massively parallel application codes on the Titan supercomputer, a turbulent combustion flow solver (S3D) and a molecular dynamics code (LAMMPS). Benchmark studies show a significant deviation from ideal weak scaling and variability in performance. The inter-task communication distance was determined to be one of the significant contributors to the performance degradation and variability. A genetic algorithm-based parallel optimization technique was used to optimize the task ordering. This technique provides an improved placement of the tasks on the nodes, taking into account the application's communication topology and the system interconnect topology. As a result, application benchmarks after task reordering through genetic algorithm show a significant improvement in performance and reduction in variability, therefore enabling the applications to achieve better time to solution and scalability on Titan during production.« less

A parallel 3-D discrete wavelet transform architecture using pipelined lifting scheme approach for video coding

NASA Astrophysics Data System (ADS)

Hegde, Ganapathi; Vaya, Pukhraj

2013-10-01

This article presents a parallel architecture for 3-D discrete wavelet transform (3-DDWT). The proposed design is based on the 1-D pipelined lifting scheme. The architecture is fully scalable beyond the present coherent Daubechies filter bank (9, 7). This 3-DDWT architecture has advantages such as no group of pictures restriction and reduced memory referencing. It offers low power consumption, low latency and high throughput. The computing technique is based on the concept that lifting scheme minimises the storage requirement. The application specific integrated circuit implementation of the proposed architecture is done by synthesising it using 65 nm Taiwan Semiconductor Manufacturing Company standard cell library. It offers a speed of 486 MHz with a power consumption of 2.56 mW. This architecture is suitable for real-time video compression even with large frame dimensions.

Gilgamesh: A Multithreaded Processor-In-Memory Architecture for Petaflops Computing

NASA Technical Reports Server (NTRS)

Sterling, T. L.; Zima, H. P.

2002-01-01

Processor-in-Memory (PIM) architectures avoid the von Neumann bottleneck in conventional machines by integrating high-density DRAM and CMOS logic on the same chip. Parallel systems based on this new technology are expected to provide higher scalability, adaptability, robustness, fault tolerance and lower power consumption than current MPPs or commodity clusters. In this paper we describe the design of Gilgamesh, a PIM-based massively parallel architecture, and elements of its execution model. Gilgamesh extends existing PIM capabilities by incorporating advanced mechanisms for virtualizing tasks and data and providing adaptive resource management for load balancing and latency tolerance. The Gilgamesh execution model is based on macroservers, a middleware layer which supports object-based runtime management of data and threads allowing explicit and dynamic control of locality and load balancing. The paper concludes with a discussion of related research activities and an outlook to future work.

A distributed-memory approximation algorithm for maximum weight perfect bipartite matching

DOE Office of Scientific and Technical Information (OSTI.GOV)

Azad, Ariful; Buluc, Aydin; Li, Xiaoye S.

We design and implement an efficient parallel approximation algorithm for the problem of maximum weight perfect matching in bipartite graphs, i.e. the problem of finding a set of non-adjacent edges that covers all vertices and has maximum weight. This problem differs from the maximum weight matching problem, for which scalable approximation algorithms are known. It is primarily motivated by finding good pivots in scalable sparse direct solvers before factorization where sequential implementations of maximum weight perfect matching algorithms, such as those available in MC64, are widely used due to the lack of scalable alternatives. To overcome this limitation, we proposemore » a fully parallel distributed memory algorithm that first generates a perfect matching and then searches for weightaugmenting cycles of length four in parallel and iteratively augments the matching with a vertex disjoint set of such cycles. For most practical problems the weights of the perfect matchings generated by our algorithm are very close to the optimum. An efficient implementation of the algorithm scales up to 256 nodes (17,408 cores) on a Cray XC40 supercomputer and can solve instances that are too large to be handled by a single node using the sequential algorithm.« less

Fan-out Estimation in Spin-based Quantum Computer Scale-up.

PubMed

Nguyen, Thien; Hill, Charles D; Hollenberg, Lloyd C L; James, Matthew R

2017-10-17

Solid-state spin-based qubits offer good prospects for scaling based on their long coherence times and nexus to large-scale electronic scale-up technologies. However, high-threshold quantum error correction requires a two-dimensional qubit array operating in parallel, posing significant challenges in fabrication and control. While architectures incorporating distributed quantum control meet this challenge head-on, most designs rely on individual control and readout of all qubits with high gate densities. We analysed the fan-out routing overhead of a dedicated control line architecture, basing the analysis on a generalised solid-state spin qubit platform parameterised to encompass Coulomb confined (e.g. donor based spin qubits) or electrostatically confined (e.g. quantum dot based spin qubits) implementations. The spatial scalability under this model is estimated using standard electronic routing methods and present-day fabrication constraints. Based on reasonable assumptions for qubit control and readout we estimate 10 2 -10 5 physical qubits, depending on the quantum interconnect implementation, can be integrated and fanned-out independently. Assuming relatively long control-free interconnects the scalability can be extended. Ultimately, the universal quantum computation may necessitate a much higher number of integrated qubits, indicating that higher dimensional electronics fabrication and/or multiplexed distributed control and readout schemes may be the preferredstrategy for large-scale implementation.

A New Conflict Resolution Method for Multiple Mobile Robots in Cluttered Environments With Motion-Liveness.

PubMed

Shahriari, Mohammadali; Biglarbegian, Mohammad

2018-01-01

This paper presents a new conflict resolution methodology for multiple mobile robots while ensuring their motion-liveness, especially for cluttered and dynamic environments. Our method constructs a mathematical formulation in a form of an optimization problem by minimizing the overall travel times of the robots subject to resolving all the conflicts in their motion. This optimization problem can be easily solved through coordinating only the robots' speeds. To overcome the computational cost in executing the algorithm for very cluttered environments, we develop an innovative method through clustering the environment into independent subproblems that can be solved using parallel programming techniques. We demonstrate the scalability of our approach through performing extensive simulations. Simulation results showed that our proposed method is capable of resolving the conflicts of 100 robots in less than 1.23 s in a cluttered environment that has 4357 intersections in the paths of the robots. We also developed an experimental testbed and demonstrated that our approach can be implemented in real time. We finally compared our approach with other existing methods in the literature both quantitatively and qualitatively. This comparison shows while our approach is mathematically sound, it is more computationally efficient, scalable for very large number of robots, and guarantees the live and smooth motion of robots.

Improving NASA's Multiscale Modeling Framework for Tropical Cyclone Climate Study

NASA Technical Reports Server (NTRS)

Shen, Bo-Wen; Nelson, Bron; Cheung, Samson; Tao, Wei-Kuo

2013-01-01

One of the current challenges in tropical cyclone (TC) research is how to improve our understanding of TC interannual variability and the impact of climate change on TCs. Recent advances in global modeling, visualization, and supercomputing technologies at NASA show potential for such studies. In this article, the authors discuss recent scalability improvement to the multiscale modeling framework (MMF) that makes it feasible to perform long-term TC-resolving simulations. The MMF consists of the finite-volume general circulation model (fvGCM), supplemented by a copy of the Goddard cumulus ensemble model (GCE) at each of the fvGCM grid points, giving 13,104 GCE copies. The original fvGCM implementation has a 1D data decomposition; the revised MMF implementation retains the 1D decomposition for most of the code, but uses a 2D decomposition for the massive copies of GCEs. Because the vast majority of computation time in the MMF is spent computing the GCEs, this approach can achieve excellent speedup without incurring the cost of modifying the entire code. Intelligent process mapping allows differing numbers of processes to be assigned to each domain for load balancing. The revised parallel implementation shows highly promising scalability, obtaining a nearly 80-fold speedup by increasing the number of cores from 30 to 3,335.

Testing New Programming Paradigms with NAS Parallel Benchmarks

NASA Technical Reports Server (NTRS)

Jin, H.; Frumkin, M.; Schultz, M.; Yan, J.

2000-01-01

Over the past decade, high performance computing has evolved rapidly, not only in hardware architectures but also with increasing complexity of real applications. Technologies have been developing to aim at scaling up to thousands of processors on both distributed and shared memory systems. Development of parallel programs on these computers is always a challenging task. Today, writing parallel programs with message passing (e.g. MPI) is the most popular way of achieving scalability and high performance. However, writing message passing programs is difficult and error prone. Recent years new effort has been made in defining new parallel programming paradigms. The best examples are: HPF (based on data parallelism) and OpenMP (based on shared memory parallelism). Both provide simple and clear extensions to sequential programs, thus greatly simplify the tedious tasks encountered in writing message passing programs. HPF is independent of memory hierarchy, however, due to the immaturity of compiler technology its performance is still questionable. Although use of parallel compiler directives is not new, OpenMP offers a portable solution in the shared-memory domain. Another important development involves the tremendous progress in the internet and its associated technology. Although still in its infancy, Java promisses portability in a heterogeneous environment and offers possibility to "compile once and run anywhere." In light of testing these new technologies, we implemented new parallel versions of the NAS Parallel Benchmarks (NPBs) with HPF and OpenMP directives, and extended the work with Java and Java-threads. The purpose of this study is to examine the effectiveness of alternative programming paradigms. NPBs consist of five kernels and three simulated applications that mimic the computation and data movement of large scale computational fluid dynamics (CFD) applications. We started with the serial version included in NPB2.3. Optimization of memory and cache usage was applied to several benchmarks, noticeably BT and SP, resulting in better sequential performance. In order to overcome the lack of an HPF performance model and guide the development of the HPF codes, we employed an empirical performance model for several primitives found in the benchmarks. We encountered a few limitations of HPF, such as lack of supporting the "REDISTRIBUTION" directive and no easy way to handle irregular computation. The parallelization with OpenMP directives was done at the outer-most loop level to achieve the largest granularity. The performance of six HPF and OpenMP benchmarks is compared with their MPI counterparts for the Class-A problem size in the figure in next page. These results were obtained on an SGI Origin2000 (195MHz) with MIPSpro-f77 compiler 7.2.1 for OpenMP and MPI codes and PGI pghpf-2.4.3 compiler with MPI interface for HPF programs.

Ray Meta: scalable de novo metagenome assembly and profiling

PubMed Central

2012-01-01

Voluminous parallel sequencing datasets, especially metagenomic experiments, require distributed computing for de novo assembly and taxonomic profiling. Ray Meta is a massively distributed metagenome assembler that is coupled with Ray Communities, which profiles microbiomes based on uniquely-colored k-mers. It can accurately assemble and profile a three billion read metagenomic experiment representing 1,000 bacterial genomes of uneven proportions in 15 hours with 1,024 processor cores, using only 1.5 GB per core. The software will facilitate the processing of large and complex datasets, and will help in generating biological insights for specific environments. Ray Meta is open source and available at http://denovoassembler.sf.net. PMID:23259615

Symplectic multi-particle tracking on GPUs

NASA Astrophysics Data System (ADS)

Liu, Zhicong; Qiang, Ji

2018-05-01

A symplectic multi-particle tracking model is implemented on the Graphic Processing Units (GPUs) using the Compute Unified Device Architecture (CUDA) language. The symplectic tracking model can preserve phase space structure and reduce non-physical effects in long term simulation, which is important for beam property evaluation in particle accelerators. Though this model is computationally expensive, it is very suitable for parallelization and can be accelerated significantly by using GPUs. In this paper, we optimized the implementation of the symplectic tracking model on both single GPU and multiple GPUs. Using a single GPU processor, the code achieves a factor of 2-10 speedup for a range of problem sizes compared with the time on a single state-of-the-art Central Processing Unit (CPU) node with similar power consumption and semiconductor technology. It also shows good scalability on a multi-GPU cluster at Oak Ridge Leadership Computing Facility. In an application to beam dynamics simulation, the GPU implementation helps save more than a factor of two total computing time in comparison to the CPU implementation.

A new tool called DISSECT for analysing large genomic data sets using a Big Data approach

PubMed Central

Canela-Xandri, Oriol; Law, Andy; Gray, Alan; Woolliams, John A.; Tenesa, Albert

2015-01-01

Large-scale genetic and genomic data are increasingly available and the major bottleneck in their analysis is a lack of sufficiently scalable computational tools. To address this problem in the context of complex traits analysis, we present DISSECT. DISSECT is a new and freely available software that is able to exploit the distributed-memory parallel computational architectures of compute clusters, to perform a wide range of genomic and epidemiologic analyses, which currently can only be carried out on reduced sample sizes or under restricted conditions. We demonstrate the usefulness of our new tool by addressing the challenge of predicting phenotypes from genotype data in human populations using mixed-linear model analysis. We analyse simulated traits from 470,000 individuals genotyped for 590,004 SNPs in ∼4 h using the combined computational power of 8,400 processor cores. We find that prediction accuracies in excess of 80% of the theoretical maximum could be achieved with large sample sizes. PMID:26657010

Use Computer-Aided Tools to Parallelize Large CFD Applications

NASA Technical Reports Server (NTRS)

Jin, H.; Frumkin, M.; Yan, J.

2000-01-01

Porting applications to high performance parallel computers is always a challenging task. It is time consuming and costly. With rapid progressing in hardware architectures and increasing complexity of real applications in recent years, the problem becomes even more sever. Today, scalability and high performance are mostly involving handwritten parallel programs using message-passing libraries (e.g. MPI). However, this process is very difficult and often error-prone. The recent reemergence of shared memory parallel (SMP) architectures, such as the cache coherent Non-Uniform Memory Access (ccNUMA) architecture used in the SGI Origin 2000, show good prospects for scaling beyond hundreds of processors. Programming on an SMP is simplified by working in a globally accessible address space. The user can supply compiler directives, such as OpenMP, to parallelize the code. As an industry standard for portable implementation of parallel programs for SMPs, OpenMP is a set of compiler directives and callable runtime library routines that extend Fortran, C and C++ to express shared memory parallelism. It promises an incremental path for parallel conversion of existing software, as well as scalability and performance for a complete rewrite or an entirely new development. Perhaps the main disadvantage of programming with directives is that inserted directives may not necessarily enhance performance. In the worst cases, it can create erroneous results. While vendors have provided tools to perform error-checking and profiling, automation in directive insertion is very limited and often failed on large programs, primarily due to the lack of a thorough enough data dependence analysis. To overcome the deficiency, we have developed a toolkit, CAPO, to automatically insert OpenMP directives in Fortran programs and apply certain degrees of optimization. CAPO is aimed at taking advantage of detailed inter-procedural dependence analysis provided by CAPTools, developed by the University of Greenwich, to reduce potential errors made by users. Earlier tests on NAS Benchmarks and ARC3D have demonstrated good success of this tool. In this study, we have applied CAPO to parallelize three large applications in the area of computational fluid dynamics (CFD): OVERFLOW, TLNS3D and INS3D. These codes are widely used for solving Navier-Stokes equations with complicated boundary conditions and turbulence model in multiple zones. Each one comprises of from 50K to 1,00k lines of FORTRAN77. As an example, CAPO took 77 hours to complete the data dependence analysis of OVERFLOW on a workstation (SGI, 175MHz, R10K processor). A fair amount of effort was spent on correcting false dependencies due to lack of necessary knowledge during the analysis. Even so, CAPO provides an easy way for user to interact with the parallelization process. The OpenMP version was generated within a day after the analysis was completed. Due to sequential algorithms involved, code sections in TLNS3D and INS3D need to be restructured by hand to produce more efficient parallel codes. An included figure shows preliminary test results of the generated OVERFLOW with several test cases in single zone. The MPI data points for the small test case were taken from a handcoded MPI version. As we can see, CAPO's version has achieved 18 fold speed up on 32 nodes of the SGI O2K. For the small test case, it outperformed the MPI version. These results are very encouraging, but further work is needed. For example, although CAPO attempts to place directives on the outer- most parallel loops in an interprocedural framework, it does not insert directives based on the best manual strategy. In particular, it lacks the support of parallelization at the multi-zone level. Future work will emphasize on the development of methodology to work in a multi-zone level and with a hybrid approach. Development of tools to perform more complicated code transformation is also needed.

Highly scalable parallel processing of extracellular recordings of Multielectrode Arrays.

PubMed

Gehring, Tiago V; Vasilaki, Eleni; Giugliano, Michele

2015-01-01

Technological advances of Multielectrode Arrays (MEAs) used for multisite, parallel electrophysiological recordings, lead to an ever increasing amount of raw data being generated. Arrays with hundreds up to a few thousands of electrodes are slowly seeing widespread use and the expectation is that more sophisticated arrays will become available in the near future. In order to process the large data volumes resulting from MEA recordings there is a pressing need for new software tools able to process many data channels in parallel. Here we present a new tool for processing MEA data recordings that makes use of new programming paradigms and recent technology developments to unleash the power of modern highly parallel hardware, such as multi-core CPUs with vector instruction sets or GPGPUs. Our tool builds on and complements existing MEA data analysis packages. It shows high scalability and can be used to speed up some performance critical pre-processing steps such as data filtering and spike detection, helping to make the analysis of larger data sets tractable.

Evaluating the performance of parallel subsurface simulators: An illustrative example with PFLOTRAN

PubMed Central

Hammond, G E; Lichtner, P C; Mills, R T

2014-01-01

[1] To better inform the subsurface scientist on the expected performance of parallel simulators, this work investigates performance of the reactive multiphase flow and multicomponent biogeochemical transport code PFLOTRAN as it is applied to several realistic modeling scenarios run on the Jaguar supercomputer. After a brief introduction to the code's parallel layout and code design, PFLOTRAN's parallel performance (measured through strong and weak scalability analyses) is evaluated in the context of conceptual model layout, software and algorithmic design, and known hardware limitations. PFLOTRAN scales well (with regard to strong scaling) for three realistic problem scenarios: (1) in situ leaching of copper from a mineral ore deposit within a 5-spot flow regime, (2) transient flow and solute transport within a regional doublet, and (3) a real-world problem involving uranium surface complexation within a heterogeneous and extremely dynamic variably saturated flow field. Weak scalability is discussed in detail for the regional doublet problem, and several difficulties with its interpretation are noted. PMID:25506097

Evaluating the performance of parallel subsurface simulators: An illustrative example with PFLOTRAN.

PubMed

Hammond, G E; Lichtner, P C; Mills, R T

2014-01-01

[1] To better inform the subsurface scientist on the expected performance of parallel simulators, this work investigates performance of the reactive multiphase flow and multicomponent biogeochemical transport code PFLOTRAN as it is applied to several realistic modeling scenarios run on the Jaguar supercomputer. After a brief introduction to the code's parallel layout and code design, PFLOTRAN's parallel performance (measured through strong and weak scalability analyses) is evaluated in the context of conceptual model layout, software and algorithmic design, and known hardware limitations. PFLOTRAN scales well (with regard to strong scaling) for three realistic problem scenarios: (1) in situ leaching of copper from a mineral ore deposit within a 5-spot flow regime, (2) transient flow and solute transport within a regional doublet, and (3) a real-world problem involving uranium surface complexation within a heterogeneous and extremely dynamic variably saturated flow field. Weak scalability is discussed in detail for the regional doublet problem, and several difficulties with its interpretation are noted.

Mining algorithm for association rules in big data based on Hadoop

NASA Astrophysics Data System (ADS)

Fu, Chunhua; Wang, Xiaojing; Zhang, Lijun; Qiao, Liying

2018-04-01

In order to solve the problem that the traditional association rules mining algorithm has been unable to meet the mining needs of large amount of data in the aspect of efficiency and scalability, take FP-Growth as an example, the algorithm is realized in the parallelization based on Hadoop framework and Map Reduce model. On the basis, it is improved using the transaction reduce method for further enhancement of the algorithm's mining efficiency. The experiment, which consists of verification of parallel mining results, comparison on efficiency between serials and parallel, variable relationship between mining time and node number and between mining time and data amount, is carried out in the mining results and efficiency by Hadoop clustering. Experiments show that the paralleled FP-Growth algorithm implemented is able to accurately mine frequent item sets, with a better performance and scalability. It can be better to meet the requirements of big data mining and efficiently mine frequent item sets and association rules from large dataset.

Scalable Quantum Networks for Distributed Computing and Sensing

DTIC Science & Technology

2016-04-01

probabilistic measurement , so we developed quantum memories and guided-wave implementations of same, demonstrating controlled delay of a heralded single...Second, fundamental scalability requires a method to synchronize protocols based on quantum measurements , which are inherently probabilistic. To meet...AFRL-AFOSR-UK-TR-2016-0007 Scalable Quantum Networks for Distributed Computing and Sensing Ian Walmsley THE UNIVERSITY OF OXFORD Final Report 04/01

A multithreaded and GPU-optimized compact finite difference algorithm for turbulent mixing at high Schmidt number using petascale computing

NASA Astrophysics Data System (ADS)

Clay, M. P.; Yeung, P. K.; Buaria, D.; Gotoh, T.

2017-11-01

Turbulent mixing at high Schmidt number is a multiscale problem which places demanding requirements on direct numerical simulations to resolve fluctuations down the to Batchelor scale. We use a dual-grid, dual-scheme and dual-communicator approach where velocity and scalar fields are computed by separate groups of parallel processes, the latter using a combined compact finite difference (CCD) scheme on finer grid with a static 3-D domain decomposition free of the communication overhead of memory transposes. A high degree of scalability is achieved for a 81923 scalar field at Schmidt number 512 in turbulence with a modest inertial range, by overlapping communication with computation whenever possible. On the Cray XE6 partition of Blue Waters, use of a dedicated thread for communication combined with OpenMP locks and nested parallelism reduces CCD timings by 34% compared to an MPI baseline. The code has been further optimized for the 27-petaflops Cray XK7 machine Titan using GPUs as accelerators with the latest OpenMP 4.5 directives, giving 2.7X speedup compared to CPU-only execution at the largest problem size. Supported by NSF Grant ACI-1036170, the NCSA Blue Waters Project with subaward via UIUC, and a DOE INCITE allocation at ORNL.

Enabling parallel simulation of large-scale HPC network systems

DOE PAGES

Mubarak, Misbah; Carothers, Christopher D.; Ross, Robert B.; ...

2016-04-07

Here, with the increasing complexity of today’s high-performance computing (HPC) architectures, simulation has become an indispensable tool for exploring the design space of HPC systems—in particular, networks. In order to make effective design decisions, simulations of these systems must possess the following properties: (1) have high accuracy and fidelity, (2) produce results in a timely manner, and (3) be able to analyze a broad range of network workloads. Most state-of-the-art HPC network simulation frameworks, however, are constrained in one or more of these areas. In this work, we present a simulation framework for modeling two important classes of networks usedmore » in today’s IBM and Cray supercomputers: torus and dragonfly networks. We use the Co-Design of Multi-layer Exascale Storage Architecture (CODES) simulation framework to simulate these network topologies at a flit-level detail using the Rensselaer Optimistic Simulation System (ROSS) for parallel discrete-event simulation. Our simulation framework meets all the requirements of a practical network simulation and can assist network designers in design space exploration. First, it uses validated and detailed flit-level network models to provide an accurate and high-fidelity network simulation. Second, instead of relying on serial time-stepped or traditional conservative discrete-event simulations that limit simulation scalability and efficiency, we use the optimistic event-scheduling capability of ROSS to achieve efficient and scalable HPC network simulations on today’s high-performance cluster systems. Third, our models give network designers a choice in simulating a broad range of network workloads, including HPC application workloads using detailed network traces, an ability that is rarely offered in parallel with high-fidelity network simulations« less

Enabling parallel simulation of large-scale HPC network systems

DOE Office of Scientific and Technical Information (OSTI.GOV)

Mubarak, Misbah; Carothers, Christopher D.; Ross, Robert B.

Here, with the increasing complexity of today’s high-performance computing (HPC) architectures, simulation has become an indispensable tool for exploring the design space of HPC systems—in particular, networks. In order to make effective design decisions, simulations of these systems must possess the following properties: (1) have high accuracy and fidelity, (2) produce results in a timely manner, and (3) be able to analyze a broad range of network workloads. Most state-of-the-art HPC network simulation frameworks, however, are constrained in one or more of these areas. In this work, we present a simulation framework for modeling two important classes of networks usedmore » in today’s IBM and Cray supercomputers: torus and dragonfly networks. We use the Co-Design of Multi-layer Exascale Storage Architecture (CODES) simulation framework to simulate these network topologies at a flit-level detail using the Rensselaer Optimistic Simulation System (ROSS) for parallel discrete-event simulation. Our simulation framework meets all the requirements of a practical network simulation and can assist network designers in design space exploration. First, it uses validated and detailed flit-level network models to provide an accurate and high-fidelity network simulation. Second, instead of relying on serial time-stepped or traditional conservative discrete-event simulations that limit simulation scalability and efficiency, we use the optimistic event-scheduling capability of ROSS to achieve efficient and scalable HPC network simulations on today’s high-performance cluster systems. Third, our models give network designers a choice in simulating a broad range of network workloads, including HPC application workloads using detailed network traces, an ability that is rarely offered in parallel with high-fidelity network simulations« less

Parallel particle filters for online identification of mechanistic mathematical models of physiology from monitoring data: performance and real-time scalability in simulation scenarios.

PubMed

Zenker, Sven

2010-08-01

Combining mechanistic mathematical models of physiology with quantitative observations using probabilistic inference may offer advantages over established approaches to computerized decision support in acute care medicine. Particle filters (PF) can perform such inference successively as data becomes available. The potential of PF for real-time state estimation (SE) for a model of cardiovascular physiology is explored using parallel computers and the ability to achieve joint state and parameter estimation (JSPE) given minimal prior knowledge tested. A parallelized sequential importance sampling/resampling algorithm was implemented and its scalability for the pure SE problem for a non-linear five-dimensional ODE model of the cardiovascular system evaluated on a Cray XT3 using up to 1,024 cores. JSPE was implemented using a state augmentation approach with artificial stochastic evolution of the parameters. Its performance when simultaneously estimating the 5 states and 18 unknown parameters when given observations only of arterial pressure, central venous pressure, heart rate, and, optionally, cardiac output, was evaluated in a simulated bleeding/resuscitation scenario. SE was successful and scaled up to 1,024 cores with appropriate algorithm parametrization, with real-time equivalent performance for up to 10 million particles. JSPE in the described underdetermined scenario achieved excellent reproduction of observables and qualitative tracking of enddiastolic ventricular volumes and sympathetic nervous activity. However, only a subset of the posterior distributions of parameters concentrated around the true values for parts of the estimated trajectories. Parallelized PF's performance makes their application to complex mathematical models of physiology for the purpose of clinical data interpretation, prediction, and therapy optimization appear promising. JSPE in the described extremely underdetermined scenario nevertheless extracted information of potential clinical relevance from the data in this simulation setting. However, fully satisfactory resolution of this problem when minimal prior knowledge about parameter values is available will require further methodological improvements, which are discussed.

Kokkos: Enabling manycore performance portability through polymorphic memory access patterns

DOE PAGES

Carter Edwards, H.; Trott, Christian R.; Sunderland, Daniel

2014-07-22

The manycore revolution can be characterized by increasing thread counts, decreasing memory per thread, and diversity of continually evolving manycore architectures. High performance computing (HPC) applications and libraries must exploit increasingly finer levels of parallelism within their codes to sustain scalability on these devices. We found that a major obstacle to performance portability is the diverse and conflicting set of constraints on memory access patterns across devices. Contemporary portable programming models address manycore parallelism (e.g., OpenMP, OpenACC, OpenCL) but fail to address memory access patterns. The Kokkos C++ library enables applications and domain libraries to achieve performance portability on diversemore » manycore architectures by unifying abstractions for both fine-grain data parallelism and memory access patterns. In this paper we describe Kokkos’ abstractions, summarize its application programmer interface (API), present performance results for unit-test kernels and mini-applications, and outline an incremental strategy for migrating legacy C++ codes to Kokkos. Furthermore, the Kokkos library is under active research and development to incorporate capabilities from new generations of manycore architectures, and to address a growing list of applications and domain libraries.« less

SQDFT: Spectral Quadrature method for large-scale parallel O(N) Kohn-Sham calculations at high temperature

NASA Astrophysics Data System (ADS)

Suryanarayana, Phanish; Pratapa, Phanisri P.; Sharma, Abhiraj; Pask, John E.

2018-03-01

We present SQDFT: a large-scale parallel implementation of the Spectral Quadrature (SQ) method for O(N) Kohn-Sham Density Functional Theory (DFT) calculations at high temperature. Specifically, we develop an efficient and scalable finite-difference implementation of the infinite-cell Clenshaw-Curtis SQ approach, in which results for the infinite crystal are obtained by expressing quantities of interest as bilinear forms or sums of bilinear forms, that are then approximated by spatially localized Clenshaw-Curtis quadrature rules. We demonstrate the accuracy of SQDFT by showing systematic convergence of energies and atomic forces with respect to SQ parameters to reference diagonalization results, and convergence with discretization to established planewave results, for both metallic and insulating systems. We further demonstrate that SQDFT achieves excellent strong and weak parallel scaling on computer systems consisting of tens of thousands of processors, with near perfect O(N) scaling with system size and wall times as low as a few seconds per self-consistent field iteration. Finally, we verify the accuracy of SQDFT in large-scale quantum molecular dynamics simulations of aluminum at high temperature.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Wylie, Brian Neil; Moreland, Kenneth D.

Graphs are a vital way of organizing data with complex correlations. A good visualization of a graph can fundamentally change human understanding of the data. Consequently, there is a rich body of work on graph visualization. Although there are many techniques that are effective on small to medium sized graphs (tens of thousands of nodes), there is a void in the research for visualizing massive graphs containing millions of nodes. Sandia is one of the few entities in the world that has the means and motivation to handle data on such a massive scale. For example, homeland security generates graphsmore » from prolific media sources such as television, telephone, and the Internet. The purpose of this project is to provide the groundwork for visualizing such massive graphs. The research provides for two major feature gaps: a parallel, interactive visualization framework and scalable algorithms to make the framework usable to a practical application. Both the frameworks and algorithms are designed to run on distributed parallel computers, which are already available at Sandia. Some features are integrated into the ThreatView{trademark} application and future work will integrate further parallel algorithms.« less

Summer Proceedings 2016: The Center for Computing Research at Sandia National Laboratories

DOE Office of Scientific and Technical Information (OSTI.GOV)

Carleton, James Brian; Parks, Michael L.

Solving sparse linear systems from the discretization of elliptic partial differential equations (PDEs) is an important building block in many engineering applications. Sparse direct solvers can solve general linear systems, but are usually slower and use much more memory than effective iterative solvers. To overcome these two disadvantages, a hierarchical solver (LoRaSp) based on H2-matrices was introduced in [22]. Here, we have developed a parallel version of the algorithm in LoRaSp to solve large sparse matrices on distributed memory machines. On a single processor, the factorization time of our parallel solver scales almost linearly with the problem size for three-dimensionalmore » problems, as opposed to the quadratic scalability of many existing sparse direct solvers. Moreover, our solver leads to almost constant numbers of iterations, when used as a preconditioner for Poisson problems. On more than one processor, our algorithm has significant speedups compared to sequential runs. With this parallel algorithm, we are able to solve large problems much faster than many existing packages as demonstrated by the numerical experiments.« less

GPU-accelerated Tersoff potentials for massively parallel Molecular Dynamics simulations

NASA Astrophysics Data System (ADS)

Nguyen, Trung Dac

2017-03-01

The Tersoff potential is one of the empirical many-body potentials that has been widely used in simulation studies at atomic scales. Unlike pair-wise potentials, the Tersoff potential involves three-body terms, which require much more arithmetic operations and data dependency. In this contribution, we have implemented the GPU-accelerated version of several variants of the Tersoff potential for LAMMPS, an open-source massively parallel Molecular Dynamics code. Compared to the existing MPI implementation in LAMMPS, the GPU implementation exhibits a better scalability and offers a speedup of 2.2X when run on 1000 compute nodes on the Titan supercomputer. On a single node, the speedup ranges from 2.0 to 8.0 times, depending on the number of atoms per GPU and hardware configurations. The most notable features of our GPU-accelerated version include its design for MPI/accelerator heterogeneous parallelism, its compatibility with other functionalities in LAMMPS, its ability to give deterministic results and to support both NVIDIA CUDA- and OpenCL-enabled accelerators. Our implementation is now part of the GPU package in LAMMPS and accessible for public use.

Scalable software-defined optical networking with high-performance routing and wavelength assignment algorithms.

PubMed

Lee, Chankyun; Cao, Xiaoyuan; Yoshikane, Noboru; Tsuritani, Takehiro; Rhee, June-Koo Kevin

2015-10-19

The feasibility of software-defined optical networking (SDON) for a practical application critically depends on scalability of centralized control performance. The paper, highly scalable routing and wavelength assignment (RWA) algorithms are investigated on an OpenFlow-based SDON testbed for proof-of-concept demonstration. Efficient RWA algorithms are proposed to achieve high performance in achieving network capacity with reduced computation cost, which is a significant attribute in a scalable centralized-control SDON. The proposed heuristic RWA algorithms differ in the orders of request processes and in the procedures of routing table updates. Combined in a shortest-path-based routing algorithm, a hottest-request-first processing policy that considers demand intensity and end-to-end distance information offers both the highest throughput of networks and acceptable computation scalability. We further investigate trade-off relationship between network throughput and computation complexity in routing table update procedure by a simulation study.

BigDebug: Debugging Primitives for Interactive Big Data Processing in Spark

PubMed Central

Gulzar, Muhammad Ali; Interlandi, Matteo; Yoo, Seunghyun; Tetali, Sai Deep; Condie, Tyson; Millstein, Todd; Kim, Miryung

2016-01-01

Developers use cloud computing platforms to process a large quantity of data in parallel when developing big data analytics. Debugging the massive parallel computations that run in today’s data-centers is time consuming and error-prone. To address this challenge, we design a set of interactive, real-time debugging primitives for big data processing in Apache Spark, the next generation data-intensive scalable cloud computing platform. This requires re-thinking the notion of step-through debugging in a traditional debugger such as gdb, because pausing the entire computation across distributed worker nodes causes significant delay and naively inspecting millions of records using a watchpoint is too time consuming for an end user. First, BIGDEBUG’s simulated breakpoints and on-demand watchpoints allow users to selectively examine distributed, intermediate data on the cloud with little overhead. Second, a user can also pinpoint a crash-inducing record and selectively resume relevant sub-computations after a quick fix. Third, a user can determine the root causes of errors (or delays) at the level of individual records through a fine-grained data provenance capability. Our evaluation shows that BIGDEBUG scales to terabytes and its record-level tracing incurs less than 25% overhead on average. It determines crash culprits orders of magnitude more accurately and provides up to 100% time saving compared to the baseline replay debugger. The results show that BIGDEBUG supports debugging at interactive speeds with minimal performance impact. PMID:27390389

STORMSeq: an open-source, user-friendly pipeline for processing personal genomics data in the cloud.

PubMed

Karczewski, Konrad J; Fernald, Guy Haskin; Martin, Alicia R; Snyder, Michael; Tatonetti, Nicholas P; Dudley, Joel T

2014-01-01

The increasing public availability of personal complete genome sequencing data has ushered in an era of democratized genomics. However, read mapping and variant calling software is constantly improving and individuals with personal genomic data may prefer to customize and update their variant calls. Here, we describe STORMSeq (Scalable Tools for Open-Source Read Mapping), a graphical interface cloud computing solution that does not require a parallel computing environment or extensive technical experience. This customizable and modular system performs read mapping, read cleaning, and variant calling and annotation. At present, STORMSeq costs approximately $2 and 5-10 hours to process a full exome sequence and $30 and 3-8 days to process a whole genome sequence. We provide this open-access and open-source resource as a user-friendly interface in Amazon EC2.

High performance computation of radiative transfer equation using the finite element method

NASA Astrophysics Data System (ADS)

Badri, M. A.; Jolivet, P.; Rousseau, B.; Favennec, Y.

2018-05-01

This article deals with an efficient strategy for numerically simulating radiative transfer phenomena using distributed computing. The finite element method alongside the discrete ordinate method is used for spatio-angular discretization of the monochromatic steady-state radiative transfer equation in an anisotropically scattering media. Two very different methods of parallelization, angular and spatial decomposition methods, are presented. To do so, the finite element method is used in a vectorial way. A detailed comparison of scalability, performance, and efficiency on thousands of processors is established for two- and three-dimensional heterogeneous test cases. Timings show that both algorithms scale well when using proper preconditioners. It is also observed that our angular decomposition scheme outperforms our domain decomposition method. Overall, we perform numerical simulations at scales that were previously unattainable by standard radiative transfer equation solvers.

Implementation and Testing of VLBI Software Correlation at the USNO

NASA Technical Reports Server (NTRS)

Fey, Alan; Ojha, Roopesh; Boboltz, Dave; Geiger, Nicole; Kingham, Kerry; Hall, David; Gaume, Ralph; Johnston, Ken

2010-01-01

The Washington Correlator (WACO) at the U.S. Naval Observatory (USNO) is a dedicated VLBI processor based on dedicated hardware of ASIC design. The WACO is currently over 10 years old and is nearing the end of its expected lifetime. Plans for implementation and testing of software correlation at the USNO are currently being considered. The VLBI correlation process is, by its very nature, well suited to a parallelized computing environment. Commercial off-the-shelf computer hardware has advanced in processing power to the point where software correlation is now both economically and technologically feasible. The advantages of software correlation are manifold but include flexibility, scalability, and easy adaptability to changing environments and requirements. We discuss our experience with and plans for use of software correlation at USNO with emphasis on the use of the DiFX software correlator.

Distributed Fast Self-Organized Maps for Massive Spectrophotometric Data Analysis †.

PubMed

Dafonte, Carlos; Garabato, Daniel; Álvarez, Marco A; Manteiga, Minia

2018-05-03

Analyzing huge amounts of data becomes essential in the era of Big Data, where databases are populated with hundreds of Gigabytes that must be processed to extract knowledge. Hence, classical algorithms must be adapted towards distributed computing methodologies that leverage the underlying computational power of these platforms. Here, a parallel, scalable, and optimized design for self-organized maps (SOM) is proposed in order to analyze massive data gathered by the spectrophotometric sensor of the European Space Agency (ESA) Gaia spacecraft, although it could be extrapolated to other domains. The performance comparison between the sequential implementation and the distributed ones based on Apache Hadoop and Apache Spark is an important part of the work, as well as the detailed analysis of the proposed optimizations. Finally, a domain-specific visualization tool to explore astronomical SOMs is presented.

Accelerating cardiac bidomain simulations using graphics processing units.

PubMed

Neic, A; Liebmann, M; Hoetzl, E; Mitchell, L; Vigmond, E J; Haase, G; Plank, G

2012-08-01

Anatomically realistic and biophysically detailed multiscale computer models of the heart are playing an increasingly important role in advancing our understanding of integrated cardiac function in health and disease. Such detailed simulations, however, are computationally vastly demanding, which is a limiting factor for a wider adoption of in-silico modeling. While current trends in high-performance computing (HPC) hardware promise to alleviate this problem, exploiting the potential of such architectures remains challenging since strongly scalable algorithms are necessitated to reduce execution times. Alternatively, acceleration technologies such as graphics processing units (GPUs) are being considered. While the potential of GPUs has been demonstrated in various applications, benefits in the context of bidomain simulations where large sparse linear systems have to be solved in parallel with advanced numerical techniques are less clear. In this study, the feasibility of multi-GPU bidomain simulations is demonstrated by running strong scalability benchmarks using a state-of-the-art model of rabbit ventricles. The model is spatially discretized using the finite element methods (FEM) on fully unstructured grids. The GPU code is directly derived from a large pre-existing code, the Cardiac Arrhythmia Research Package (CARP), with very minor perturbation of the code base. Overall, bidomain simulations were sped up by a factor of 11.8 to 16.3 in benchmarks running on 6-20 GPUs compared to the same number of CPU cores. To match the fastest GPU simulation which engaged 20 GPUs, 476 CPU cores were required on a national supercomputing facility.

Accelerating Cardiac Bidomain Simulations Using Graphics Processing Units

PubMed Central

Neic, Aurel; Liebmann, Manfred; Hoetzl, Elena; Mitchell, Lawrence; Vigmond, Edward J.; Haase, Gundolf

2013-01-01

Anatomically realistic and biophysically detailed multiscale computer models of the heart are playing an increasingly important role in advancing our understanding of integrated cardiac function in health and disease. Such detailed simulations, however, are computationally vastly demanding, which is a limiting factor for a wider adoption of in-silico modeling. While current trends in high-performance computing (HPC) hardware promise to alleviate this problem, exploiting the potential of such architectures remains challenging since strongly scalable algorithms are necessitated to reduce execution times. Alternatively, acceleration technologies such as graphics processing units (GPUs) are being considered. While the potential of GPUs has been demonstrated in various applications, benefits in the context of bidomain simulations where large sparse linear systems have to be solved in parallel with advanced numerical techniques are less clear. In this study, the feasibility of multi-GPU bidomain simulations is demonstrated by running strong scalability benchmarks using a state-of-the-art model of rabbit ventricles. The model is spatially discretized using the finite element methods (FEM) on fully unstructured grids. The GPU code is directly derived from a large pre-existing code, the Cardiac Arrhythmia Research Package (CARP), with very minor perturbation of the code base. Overall, bidomain simulations were sped up by a factor of 11.8 to 16.3 in benchmarks running on 6–20 GPUs compared to the same number of CPU cores. To match the fastest GPU simulation which engaged 20GPUs, 476 CPU cores were required on a national supercomputing facility. PMID:22692867