Accelerating epistasis analysis in human genetics with consumer graphics hardware.

PubMed

Sinnott-Armstrong, Nicholas A; Greene, Casey S; Cancare, Fabio; Moore, Jason H

2009-07-24

Human geneticists are now capable of measuring more than one million DNA sequence variations from across the human genome. The new challenge is to develop computationally feasible methods capable of analyzing these data for associations with common human disease, particularly in the context of epistasis. Epistasis describes the situation where multiple genes interact in a complex non-linear manner to determine an individual's disease risk and is thought to be ubiquitous for common diseases. Multifactor Dimensionality Reduction (MDR) is an algorithm capable of detecting epistasis. An exhaustive analysis with MDR is often computationally expensive, particularly for high order interactions. This challenge has previously been met with parallel computation and expensive hardware. The option we examine here exploits commodity hardware designed for computer graphics. In modern computers Graphics Processing Units (GPUs) have more memory bandwidth and computational capability than Central Processing Units (CPUs) and are well suited to this problem. Advances in the video game industry have led to an economy of scale creating a situation where these powerful components are readily available at very low cost. Here we implement and evaluate the performance of the MDR algorithm on GPUs. Of primary interest are the time required for an epistasis analysis and the price to performance ratio of available solutions. We found that using MDR on GPUs consistently increased performance per machine over both a feature rich Java software package and a C++ cluster implementation. The performance of a GPU workstation running a GPU implementation reduces computation time by a factor of 160 compared to an 8-core workstation running the Java implementation on CPUs. This GPU workstation performs similarly to 150 cores running an optimized C++ implementation on a Beowulf cluster. Furthermore this GPU system provides extremely cost effective performance while leaving the CPU available for other tasks. The GPU workstation containing three GPUs costs $2000 while obtaining similar performance on a Beowulf cluster requires 150 CPU cores which, including the added infrastructure and support cost of the cluster system, cost approximately $82,500. Graphics hardware based computing provides a cost effective means to perform genetic analysis of epistasis using MDR on large datasets without the infrastructure of a computing cluster.

Cognitive Model Exploration and Optimization: A New Challenge for Computational Science

DTIC Science & Technology

2010-03-01

the generation and analysis of computational cognitive models to explain various aspects of cognition. Typically the behavior of these models...computational scale of a workstation, so we have turned to high performance computing (HPC) clusters and volunteer computing for large-scale...computational resources. The majority of applications on the Department of Defense HPC clusters focus on solving partial differential equations (Post

Improved Ant Colony Clustering Algorithm and Its Performance Study

PubMed Central

Gao, Wei

2016-01-01

Clustering analysis is used in many disciplines and applications; it is an important tool that descriptively identifies homogeneous groups of objects based on attribute values. The ant colony clustering algorithm is a swarm-intelligent method used for clustering problems that is inspired by the behavior of ant colonies that cluster their corpses and sort their larvae. A new abstraction ant colony clustering algorithm using a data combination mechanism is proposed to improve the computational efficiency and accuracy of the ant colony clustering algorithm. The abstraction ant colony clustering algorithm is used to cluster benchmark problems, and its performance is compared with the ant colony clustering algorithm and other methods used in existing literature. Based on similar computational difficulties and complexities, the results show that the abstraction ant colony clustering algorithm produces results that are not only more accurate but also more efficiently determined than the ant colony clustering algorithm and the other methods. Thus, the abstraction ant colony clustering algorithm can be used for efficient multivariate data clustering. PMID:26839533

Performance comparison analysis library communication cluster system using merge sort

NASA Astrophysics Data System (ADS)

Wulandari, D. A. R.; Ramadhan, M. E.

2018-04-01

Begins by using a single processor, to increase the speed of computing time, the use of multi-processor was introduced. The second paradigm is known as parallel computing, example cluster. The cluster must have the communication potocol for processing, one of it is message passing Interface (MPI). MPI have many library, both of them OPENMPI and MPICH2. Performance of the cluster machine depend on suitable between performance characters of library communication and characters of the problem so this study aims to analyze the comparative performances libraries in handling parallel computing process. The case study in this research are MPICH2 and OpenMPI. This case research execute sorting’s problem to know the performance of cluster system. The sorting problem use mergesort method. The research method is by implementing OpenMPI and MPICH2 on a Linux-based cluster by using five computer virtual then analyze the performance of the system by different scenario tests and three parameters for to know the performance of MPICH2 and OpenMPI. These performances are execution time, speedup and efficiency. The results of this study showed that the addition of each data size makes OpenMPI and MPICH2 have an average speed-up and efficiency tend to increase but at a large data size decreases. increased data size doesn’t necessarily increased speed up and efficiency but only execution time example in 100000 data size. OpenMPI has a execution time greater than MPICH2 example in 1000 data size average execution time with MPICH2 is 0,009721 and OpenMPI is 0,003895 OpenMPI can customize communication needs.

Running climate model on a commercial cloud computing environment: A case study using Community Earth System Model (CESM) on Amazon AWS

NASA Astrophysics Data System (ADS)

Chen, Xiuhong; Huang, Xianglei; Jiao, Chaoyi; Flanner, Mark G.; Raeker, Todd; Palen, Brock

2017-01-01

The suites of numerical models used for simulating climate of our planet are usually run on dedicated high-performance computing (HPC) resources. This study investigates an alternative to the usual approach, i.e. carrying out climate model simulations on commercially available cloud computing environment. We test the performance and reliability of running the CESM (Community Earth System Model), a flagship climate model in the United States developed by the National Center for Atmospheric Research (NCAR), on Amazon Web Service (AWS) EC2, the cloud computing environment by Amazon.com, Inc. StarCluster is used to create virtual computing cluster on the AWS EC2 for the CESM simulations. The wall-clock time for one year of CESM simulation on the AWS EC2 virtual cluster is comparable to the time spent for the same simulation on a local dedicated high-performance computing cluster with InfiniBand connections. The CESM simulation can be efficiently scaled with the number of CPU cores on the AWS EC2 virtual cluster environment up to 64 cores. For the standard configuration of the CESM at a spatial resolution of 1.9° latitude by 2.5° longitude, increasing the number of cores from 16 to 64 reduces the wall-clock running time by more than 50% and the scaling is nearly linear. Beyond 64 cores, the communication latency starts to outweigh the benefit of distributed computing and the parallel speedup becomes nearly unchanged.

Optimizing R with SparkR on a commodity cluster for biomedical research.

PubMed

Sedlmayr, Martin; Würfl, Tobias; Maier, Christian; Häberle, Lothar; Fasching, Peter; Prokosch, Hans-Ulrich; Christoph, Jan

2016-12-01

Medical researchers are challenged today by the enormous amount of data collected in healthcare. Analysis methods such as genome-wide association studies (GWAS) are often computationally intensive and thus require enormous resources to be performed in a reasonable amount of time. While dedicated clusters and public clouds may deliver the desired performance, their use requires upfront financial efforts or anonymous data, which is often not possible for preliminary or occasional tasks. We explored the possibilities to build a private, flexible cluster for processing scripts in R based on commodity, non-dedicated hardware of our department. For this, a GWAS-calculation in R on a single desktop computer, a Message Passing Interface (MPI)-cluster, and a SparkR-cluster were compared with regards to the performance, scalability, quality, and simplicity. The original script had a projected runtime of three years on a single desktop computer. Optimizing the script in R already yielded a significant reduction in computing time (2 weeks). By using R-MPI and SparkR, we were able to parallelize the computation and reduce the time to less than three hours (2.6 h) on already available, standard office computers. While MPI is a proven approach in high-performance clusters, it requires rather static, dedicated nodes. SparkR and its Hadoop siblings allow for a dynamic, elastic environment with automated failure handling. SparkR also scales better with the number of nodes in the cluster than MPI due to optimized data communication. R is a popular environment for clinical data analysis. The new SparkR solution offers elastic resources and allows supporting big data analysis using R even on non-dedicated resources with minimal change to the original code. To unleash the full potential, additional efforts should be invested to customize and improve the algorithms, especially with regards to data distribution. Copyright © 2016 The Authors. Published by Elsevier Ireland Ltd.. All rights reserved.

A parallel-processing approach to computing for the geographic sciences; applications and systems enhancements

USGS Publications Warehouse

Crane, Michael; Steinwand, Dan; Beckmann, Tim; Krpan, Greg; Liu, Shu-Guang; Nichols, Erin; Haga, Jim; Maddox, Brian; Bilderback, Chris; Feller, Mark; Homer, George

2001-01-01

The overarching goal of this project is to build a spatially distributed infrastructure for information science research by forming a team of information science researchers and providing them with similar hardware and software tools to perform collaborative research. Four geographically distributed Centers of the U.S. Geological Survey (USGS) are developing their own clusters of low-cost, personal computers into parallel computing environments that provide a costeffective way for the USGS to increase participation in the high-performance computing community. Referred to as Beowulf clusters, these hybrid systems provide the robust computing power required for conducting information science research into parallel computing systems and applications.

The `TTIME' Package: Performance Evaluation in a Cluster Computing Environment

NASA Astrophysics Data System (ADS)

Howe, Marico; Berleant, Daniel; Everett, Albert

2011-06-01

The objective of translating developmental event time across mammalian species is to gain an understanding of the timing of human developmental events based on known time of those events in animals. The potential benefits include improvements to diagnostic and intervention capabilities. The CRAN `ttime' package provides the functionality to infer unknown event timings and investigate phylogenetic proximity utilizing hierarchical clustering of both known and predicted event timings. The original generic mammalian model included nine eutherian mammals: Felis domestica (cat), Mustela putorius furo (ferret), Mesocricetus auratus (hamster), Macaca mulatta (monkey), Homo sapiens (humans), Mus musculus (mouse), Oryctolagus cuniculus (rabbit), Rattus norvegicus (rat), and Acomys cahirinus (spiny mouse). However, the data for this model is expected to grow as more data about developmental events is identified and incorporated into the analysis. Performance evaluation of the `ttime' package across a cluster computing environment versus a comparative analysis in a serial computing environment provides an important computational performance assessment. A theoretical analysis is the first stage of a process in which the second stage, if justified by the theoretical analysis, is to investigate an actual implementation of the `ttime' package in a cluster computing environment and to understand the parallelization process that underlies implementation.

Large Data at Small Universities: Astronomical processing using a computer classroom

NASA Astrophysics Data System (ADS)

Fuller, Nathaniel James; Clarkson, William I.; Fluharty, Bill; Belanger, Zach; Dage, Kristen

2016-06-01

The use of large computing clusters for astronomy research is becoming more commonplace as datasets expand, but access to these required resources is sometimes difficult for research groups working at smaller Universities. As an alternative to purchasing processing time on an off-site computing cluster, or purchasing dedicated hardware, we show how one can easily build a crude on-site cluster by utilizing idle cycles on instructional computers in computer-lab classrooms. Since these computers are maintained as part of the educational mission of the University, the resource impact on the investigator is generally low.By using open source Python routines, it is possible to have a large number of desktop computers working together via a local network to sort through large data sets. By running traditional analysis routines in an “embarrassingly parallel” manner, gains in speed are accomplished without requiring the investigator to learn how to write routines using highly specialized methodology. We demonstrate this concept here applied to 1. photometry of large-format images and 2. Statistical significance-tests for X-ray lightcurve analysis. In these scenarios, we see a speed-up factor which scales almost linearly with the number of cores in the cluster. Additionally, we show that the usage of the cluster does not severely limit performance for a local user, and indeed the processing can be performed while the computers are in use for classroom purposes.

Message Passing vs. Shared Address Space on a Cluster of SMPs

NASA Technical Reports Server (NTRS)

Shan, Hongzhang; Singh, Jaswinder Pal; Oliker, Leonid; Biswas, Rupak

2000-01-01

The convergence of scalable computer architectures using clusters of PCs (or PC-SMPs) with commodity networking has become an attractive platform for high end scientific computing. Currently, message-passing and shared address space (SAS) are the two leading programming paradigms for these systems. Message-passing has been standardized with MPI, and is the most common and mature programming approach. However message-passing code development can be extremely difficult, especially for irregular 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 programming effort, required for six applications under both programming models on a 32 CPU PC-SMP cluster. Our application suite consists of codes that typically do not exhibit high efficiency under shared memory programming. due to their high communication to computation ratios and complex communication patterns. Results indicate that SAS can achieve about half the parallel efficiency of MPI for most of our applications: however, on certain classes of problems SAS performance is competitive with MPI. We also present new algorithms for improving the PC cluster performance of MPI collective operations.

High Performance Computing of Meshless Time Domain Method on Multi-GPU Cluster

NASA Astrophysics Data System (ADS)

Ikuno, Soichiro; Nakata, Susumu; Hirokawa, Yuta; Itoh, Taku

2015-01-01

High performance computing of Meshless Time Domain Method (MTDM) on multi-GPU using the supercomputer HA-PACS (Highly Accelerated Parallel Advanced system for Computational Sciences) at University of Tsukuba is investigated. Generally, the finite difference time domain (FDTD) method is adopted for the numerical simulation of the electromagnetic wave propagation phenomena. However, the numerical domain must be divided into rectangle meshes, and it is difficult to adopt the problem in a complexed domain to the method. On the other hand, MTDM can be easily adept to the problem because MTDM does not requires meshes. In the present study, we implement MTDM on multi-GPU cluster to speedup the method, and numerically investigate the performance of the method on multi-GPU cluster. To reduce the computation time, the communication time between the decomposed domain is hided below the perfect matched layer (PML) calculation procedure. The results of computation show that speedup of MTDM on 128 GPUs is 173 times faster than that of single CPU calculation.

Efficient architecture for spike sorting in reconfigurable hardware.

PubMed

Hwang, Wen-Jyi; Lee, Wei-Hao; Lin, Shiow-Jyu; Lai, Sheng-Ying

2013-11-01

This paper presents a novel hardware architecture for fast spike sorting. The architecture is able to perform both the feature extraction and clustering in hardware. The generalized Hebbian algorithm (GHA) and fuzzy C-means (FCM) algorithm are used for feature extraction and clustering, respectively. The employment of GHA allows efficient computation of principal components for subsequent clustering operations. The FCM is able to achieve near optimal clustering for spike sorting. Its performance is insensitive to the selection of initial cluster centers. The hardware implementations of GHA and FCM feature low area costs and high throughput. In the GHA architecture, the computation of different weight vectors share the same circuit for lowering the area costs. Moreover, in the FCM hardware implementation, the usual iterative operations for updating the membership matrix and cluster centroid are merged into one single updating process to evade the large storage requirement. To show the effectiveness of the circuit, the proposed architecture is physically implemented by field programmable gate array (FPGA). It is embedded in a System-on-Chip (SOC) platform for performance measurement. Experimental results show that the proposed architecture is an efficient spike sorting design for attaining high classification correct rate and high speed computation.

JMS: An Open Source Workflow Management System and Web-Based Cluster Front-End for High Performance Computing.

PubMed

Brown, David K; Penkler, David L; Musyoka, Thommas M; Bishop, Özlem Tastan

2015-01-01

Complex computational pipelines are becoming a staple of modern scientific research. Often these pipelines are resource intensive and require days of computing time. In such cases, it makes sense to run them over high performance computing (HPC) clusters where they can take advantage of the aggregated resources of many powerful computers. In addition to this, researchers often want to integrate their workflows into their own web servers. In these cases, software is needed to manage the submission of jobs from the web interface to the cluster and then return the results once the job has finished executing. We have developed the Job Management System (JMS), a workflow management system and web interface for high performance computing (HPC). JMS provides users with a user-friendly web interface for creating complex workflows with multiple stages. It integrates this workflow functionality with the resource manager, a tool that is used to control and manage batch jobs on HPC clusters. As such, JMS combines workflow management functionality with cluster administration functionality. In addition, JMS provides developer tools including a code editor and the ability to version tools and scripts. JMS can be used by researchers from any field to build and run complex computational pipelines and provides functionality to include these pipelines in external interfaces. JMS is currently being used to house a number of bioinformatics pipelines at the Research Unit in Bioinformatics (RUBi) at Rhodes University. JMS is an open-source project and is freely available at https://github.com/RUBi-ZA/JMS.

JMS: An Open Source Workflow Management System and Web-Based Cluster Front-End for High Performance Computing

PubMed Central

Brown, David K.; Penkler, David L.; Musyoka, Thommas M.; Bishop, Özlem Tastan

2015-01-01

Complex computational pipelines are becoming a staple of modern scientific research. Often these pipelines are resource intensive and require days of computing time. In such cases, it makes sense to run them over high performance computing (HPC) clusters where they can take advantage of the aggregated resources of many powerful computers. In addition to this, researchers often want to integrate their workflows into their own web servers. In these cases, software is needed to manage the submission of jobs from the web interface to the cluster and then return the results once the job has finished executing. We have developed the Job Management System (JMS), a workflow management system and web interface for high performance computing (HPC). JMS provides users with a user-friendly web interface for creating complex workflows with multiple stages. It integrates this workflow functionality with the resource manager, a tool that is used to control and manage batch jobs on HPC clusters. As such, JMS combines workflow management functionality with cluster administration functionality. In addition, JMS provides developer tools including a code editor and the ability to version tools and scripts. JMS can be used by researchers from any field to build and run complex computational pipelines and provides functionality to include these pipelines in external interfaces. JMS is currently being used to house a number of bioinformatics pipelines at the Research Unit in Bioinformatics (RUBi) at Rhodes University. JMS is an open-source project and is freely available at https://github.com/RUBi-ZA/JMS. PMID:26280450

Low cost, high performance processing of single particle cryo-electron microscopy data in the cloud.

PubMed

Cianfrocco, Michael A; Leschziner, Andres E

2015-05-08

The advent of a new generation of electron microscopes and direct electron detectors has realized the potential of single particle cryo-electron microscopy (cryo-EM) as a technique to generate high-resolution structures. Calculating these structures requires high performance computing clusters, a resource that may be limiting to many likely cryo-EM users. To address this limitation and facilitate the spread of cryo-EM, we developed a publicly available 'off-the-shelf' computing environment on Amazon's elastic cloud computing infrastructure. This environment provides users with single particle cryo-EM software packages and the ability to create computing clusters with 16-480+ CPUs. We tested our computing environment using a publicly available 80S yeast ribosome dataset and estimate that laboratories could determine high-resolution cryo-EM structures for $50 to $1500 per structure within a timeframe comparable to local clusters. Our analysis shows that Amazon's cloud computing environment may offer a viable computing environment for cryo-EM.

On efficiency of fire simulation realization: parallelization with greater number of computational meshes

NASA Astrophysics Data System (ADS)

Valasek, Lukas; Glasa, Jan

2017-12-01

Current fire simulation systems are capable to utilize advantages of high-performance computer (HPC) platforms available and to model fires efficiently in parallel. In this paper, efficiency of a corridor fire simulation on a HPC computer cluster is discussed. The parallel MPI version of Fire Dynamics Simulator is used for testing efficiency of selected strategies of allocation of computational resources of the cluster using a greater number of computational cores. Simulation results indicate that if the number of cores used is not equal to a multiple of the total number of cluster node cores there are allocation strategies which provide more efficient calculations.

Spike sorting using locality preserving projection with gap statistics and landmark-based spectral clustering.

PubMed

Nguyen, Thanh; Khosravi, Abbas; Creighton, Douglas; Nahavandi, Saeid

2014-12-30

Understanding neural functions requires knowledge from analysing electrophysiological data. The process of assigning spikes of a multichannel signal into clusters, called spike sorting, is one of the important problems in such analysis. There have been various automated spike sorting techniques with both advantages and disadvantages regarding accuracy and computational costs. Therefore, developing spike sorting methods that are highly accurate and computationally inexpensive is always a challenge in the biomedical engineering practice. An automatic unsupervised spike sorting method is proposed in this paper. The method uses features extracted by the locality preserving projection (LPP) algorithm. These features afterwards serve as inputs for the landmark-based spectral clustering (LSC) method. Gap statistics (GS) is employed to evaluate the number of clusters before the LSC can be performed. The proposed LPP-LSC is highly accurate and computationally inexpensive spike sorting approach. LPP spike features are very discriminative; thereby boost the performance of clustering methods. Furthermore, the LSC method exhibits its efficiency when integrated with the cluster evaluator GS. The proposed method's accuracy is approximately 13% superior to that of the benchmark combination between wavelet transformation and superparamagnetic clustering (WT-SPC). Additionally, LPP-LSC computing time is six times less than that of the WT-SPC. LPP-LSC obviously demonstrates a win-win spike sorting solution meeting both accuracy and computational cost criteria. LPP and LSC are linear algorithms that help reduce computational burden and thus their combination can be applied into real-time spike analysis. Copyright © 2014 Elsevier B.V. All rights reserved.

P2P Technology for High-Performance Computing: An Overview

NASA Technical Reports Server (NTRS)

Follen, Gregory J. (Technical Monitor); Berry, Jason

2003-01-01

The transition from cluster computing to peer-to-peer (P2P) high-performance computing has recently attracted the attention of the computer science community. It has been recognized that existing local networks and dedicated clusters of headless workstations can serve as inexpensive yet powerful virtual supercomputers. It has also been recognized that the vast number of lower-end computers connected to the Internet stay idle for as long as 90% of the time. The growing speed of Internet connections and the high availability of free CPU time encourage exploration of the possibility to use the whole Internet rather than local clusters as massively parallel yet almost freely available P2P supercomputer. As a part of a larger project on P2P high-performance computing, it has been my goal to compile an overview of the 2P2 paradigm. I have studied various P2P platforms and I have compiled systematic brief descriptions of their most important characteristics. I have also experimented and obtained hands-on experience with selected P2P platforms focusing on those that seem promising with respect to P2P high-performance computing. I have also compiled relevant literature and web references. I have prepared a draft technical report and I have summarized my findings in a poster paper.

TOSCA-based orchestration of complex clusters at the IaaS level

NASA Astrophysics Data System (ADS)

Caballer, M.; Donvito, G.; Moltó, G.; Rocha, R.; Velten, M.

2017-10-01

This paper describes the adoption and extension of the TOSCA standard by the INDIGO-DataCloud project for the definition and deployment of complex computing clusters together with the required support in both OpenStack and OpenNebula, carried out in close collaboration with industry partners such as IBM. Two examples of these clusters are described in this paper, the definition of an elastic computing cluster to support the Galaxy bioinformatics application where the nodes are dynamically added and removed from the cluster to adapt to the workload, and the definition of an scalable Apache Mesos cluster for the execution of batch jobs and support for long-running services. The coupling of TOSCA with Ansible Roles to perform automated installation has resulted in the definition of high-level, deterministic templates to provision complex computing clusters across different Cloud sites.

RGCA: A Reliable GPU Cluster Architecture for Large-Scale Internet of Things Computing Based on Effective Performance-Energy Optimization

PubMed Central

Chen, Qingkui; Zhao, Deyu; Wang, Jingjuan

2017-01-01

This paper aims to develop a low-cost, high-performance and high-reliability computing system to process large-scale data using common data mining algorithms in the Internet of Things (IoT) computing environment. Considering the characteristics of IoT data processing, similar to mainstream high performance computing, we use a GPU (Graphics Processing Unit) cluster to achieve better IoT services. Firstly, we present an energy consumption calculation method (ECCM) based on WSNs. Then, using the CUDA (Compute Unified Device Architecture) Programming model, we propose a Two-level Parallel Optimization Model (TLPOM) which exploits reasonable resource planning and common compiler optimization techniques to obtain the best blocks and threads configuration considering the resource constraints of each node. The key to this part is dynamic coupling Thread-Level Parallelism (TLP) and Instruction-Level Parallelism (ILP) to improve the performance of the algorithms without additional energy consumption. Finally, combining the ECCM and the TLPOM, we use the Reliable GPU Cluster Architecture (RGCA) to obtain a high-reliability computing system considering the nodes’ diversity, algorithm characteristics, etc. The results show that the performance of the algorithms significantly increased by 34.1%, 33.96% and 24.07% for Fermi, Kepler and Maxwell on average with TLPOM and the RGCA ensures that our IoT computing system provides low-cost and high-reliability services. PMID:28777325

RGCA: A Reliable GPU Cluster Architecture for Large-Scale Internet of Things Computing Based on Effective Performance-Energy Optimization.

PubMed

Fang, Yuling; Chen, Qingkui; Xiong, Neal N; Zhao, Deyu; Wang, Jingjuan

2017-08-04

This paper aims to develop a low-cost, high-performance and high-reliability computing system to process large-scale data using common data mining algorithms in the Internet of Things (IoT) computing environment. Considering the characteristics of IoT data processing, similar to mainstream high performance computing, we use a GPU (Graphics Processing Unit) cluster to achieve better IoT services. Firstly, we present an energy consumption calculation method (ECCM) based on WSNs. Then, using the CUDA (Compute Unified Device Architecture) Programming model, we propose a Two-level Parallel Optimization Model (TLPOM) which exploits reasonable resource planning and common compiler optimization techniques to obtain the best blocks and threads configuration considering the resource constraints of each node. The key to this part is dynamic coupling Thread-Level Parallelism (TLP) and Instruction-Level Parallelism (ILP) to improve the performance of the algorithms without additional energy consumption. Finally, combining the ECCM and the TLPOM, we use the Reliable GPU Cluster Architecture (RGCA) to obtain a high-reliability computing system considering the nodes' diversity, algorithm characteristics, etc. The results show that the performance of the algorithms significantly increased by 34.1%, 33.96% and 24.07% for Fermi, Kepler and Maxwell on average with TLPOM and the RGCA ensures that our IoT computing system provides low-cost and high-reliability services.

Subspace Clustering via Learning an Adaptive Low-Rank Graph.

PubMed

Yin, Ming; Xie, Shengli; Wu, Zongze; Zhang, Yun; Gao, Junbin

2018-08-01

By using a sparse representation or low-rank representation of data, the graph-based subspace clustering has recently attracted considerable attention in computer vision, given its capability and efficiency in clustering data. However, the graph weights built using the representation coefficients are not the exact ones as the traditional definition is in a deterministic way. The two steps of representation and clustering are conducted in an independent manner, thus an overall optimal result cannot be guaranteed. Furthermore, it is unclear how the clustering performance will be affected by using this graph. For example, the graph parameters, i.e., the weights on edges, have to be artificially pre-specified while it is very difficult to choose the optimum. To this end, in this paper, a novel subspace clustering via learning an adaptive low-rank graph affinity matrix is proposed, where the affinity matrix and the representation coefficients are learned in a unified framework. As such, the pre-computed graph regularizer is effectively obviated and better performance can be achieved. Experimental results on several famous databases demonstrate that the proposed method performs better against the state-of-the-art approaches, in clustering.

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

Allada, Veerendra, Benjegerdes, Troy; Bode, Brett

Commodity clusters augmented with application accelerators are evolving as competitive high performance computing systems. The Graphical Processing Unit (GPU) with a very high arithmetic density and performance per price ratio is a good platform for the scientific application acceleration. In addition to the interconnect bottlenecks among the cluster compute nodes, the cost of memory copies between the host and the GPU device have to be carefully amortized to improve the overall efficiency of the application. Scientific applications also rely on efficient implementation of the BAsic Linear Algebra Subroutines (BLAS), among which the General Matrix Multiply (GEMM) is considered as themore » workhorse subroutine. In this paper, they study the performance of the memory copies and GEMM subroutines that are critical to port the computational chemistry algorithms to the GPU clusters. To that end, a benchmark based on the NetPIPE framework is developed to evaluate the latency and bandwidth of the memory copies between the host and the GPU device. The performance of the single and double precision GEMM subroutines from the NVIDIA CUBLAS 2.0 library are studied. The results have been compared with that of the BLAS routines from the Intel Math Kernel Library (MKL) to understand the computational trade-offs. The test bed is a Intel Xeon cluster equipped with NVIDIA Tesla GPUs.« less

STEMsalabim: A high-performance computing cluster friendly code for scanning transmission electron microscopy image simulations of thin specimens.

PubMed

Oelerich, Jan Oliver; Duschek, Lennart; Belz, Jürgen; Beyer, Andreas; Baranovskii, Sergei D; Volz, Kerstin

2017-06-01

We present a new multislice code for the computer simulation of scanning transmission electron microscope (STEM) images based on the frozen lattice approximation. Unlike existing software packages, the code is optimized to perform well on highly parallelized computing clusters, combining distributed and shared memory architectures. This enables efficient calculation of large lateral scanning areas of the specimen within the frozen lattice approximation and fine-grained sweeps of parameter space. Copyright © 2017 Elsevier B.V. All rights reserved.

Parallel hyperbolic PDE simulation on clusters: Cell versus GPU

NASA Astrophysics Data System (ADS)

Rostrup, Scott; De Sterck, Hans

2010-12-01

Increasingly, high-performance computing is looking towards data-parallel computational devices to enhance computational performance. Two technologies that have received significant attention are IBM's Cell Processor and NVIDIA's CUDA programming model for graphics processing unit (GPU) computing. In this paper we investigate the acceleration of parallel hyperbolic partial differential equation simulation on structured grids with explicit time integration on clusters with Cell and GPU backends. The message passing interface (MPI) is used for communication between nodes at the coarsest level of parallelism. Optimizations of the simulation code at the several finer levels of parallelism that the data-parallel devices provide are described in terms of data layout, data flow and data-parallel instructions. Optimized Cell and GPU performance are compared with reference code performance on a single x86 central processing unit (CPU) core in single and double precision. We further compare the CPU, Cell and GPU platforms on a chip-to-chip basis, and compare performance on single cluster nodes with two CPUs, two Cell processors or two GPUs in a shared memory configuration (without MPI). We finally compare performance on clusters with 32 CPUs, 32 Cell processors, and 32 GPUs using MPI. Our GPU cluster results use NVIDIA Tesla GPUs with GT200 architecture, but some preliminary results on recently introduced NVIDIA GPUs with the next-generation Fermi architecture are also included. This paper provides computational scientists and engineers who are considering porting their codes to accelerator environments with insight into how structured grid based explicit algorithms can be optimized for clusters with Cell and GPU accelerators. It also provides insight into the speed-up that may be gained on current and future accelerator architectures for this class of applications. Program summaryProgram title: SWsolver Catalogue identifier: AEGY_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEGY_v1_0.html Program obtainable from: CPC Program Library, Queen's University, Belfast, N. Ireland Licensing provisions: GPL v3 No. of lines in distributed program, including test data, etc.: 59 168 No. of bytes in distributed program, including test data, etc.: 453 409 Distribution format: tar.gz Programming language: C, CUDA Computer: Parallel Computing Clusters. Individual compute nodes may consist of x86 CPU, Cell processor, or x86 CPU with attached NVIDIA GPU accelerator. Operating system: Linux Has the code been vectorised or parallelized?: Yes. Tested on 1-128 x86 CPU cores, 1-32 Cell Processors, and 1-32 NVIDIA GPUs. RAM: Tested on Problems requiring up to 4 GB per compute node. Classification: 12 External routines: MPI, CUDA, IBM Cell SDK Nature of problem: MPI-parallel simulation of Shallow Water equations using high-resolution 2D hyperbolic equation solver on regular Cartesian grids for x86 CPU, Cell Processor, and NVIDIA GPU using CUDA. Solution method: SWsolver provides 3 implementations of a high-resolution 2D Shallow Water equation solver on regular Cartesian grids, for CPU, Cell Processor, and NVIDIA GPU. Each implementation uses MPI to divide work across a parallel computing cluster. Additional comments: Sub-program numdiff is used for the test run.

Research on retailer data clustering algorithm based on Spark

NASA Astrophysics Data System (ADS)

Huang, Qiuman; Zhou, Feng

2017-03-01

Big data analysis is a hot topic in the IT field now. Spark is a high-reliability and high-performance distributed parallel computing framework for big data sets. K-means algorithm is one of the classical partition methods in clustering algorithm. In this paper, we study the k-means clustering algorithm on Spark. Firstly, the principle of the algorithm is analyzed, and then the clustering analysis is carried out on the supermarket customers through the experiment to find out the different shopping patterns. At the same time, this paper proposes the parallelization of k-means algorithm and the distributed computing framework of Spark, and gives the concrete design scheme and implementation scheme. This paper uses the two-year sales data of a supermarket to validate the proposed clustering algorithm and achieve the goal of subdividing customers, and then analyze the clustering results to help enterprises to take different marketing strategies for different customer groups to improve sales performance.

Bioinformatics and Astrophysics Cluster (BinAc)

NASA Astrophysics Data System (ADS)

Krüger, Jens; Lutz, Volker; Bartusch, Felix; Dilling, Werner; Gorska, Anna; Schäfer, Christoph; Walter, Thomas

2017-09-01

BinAC provides central high performance computing capacities for bioinformaticians and astrophysicists from the state of Baden-Württemberg. The bwForCluster BinAC is part of the implementation concept for scientific computing for the universities in Baden-Württemberg. Community specific support is offered through the bwHPC-C5 project.

Low cost, high performance processing of single particle cryo-electron microscopy data in the cloud

PubMed Central

Cianfrocco, Michael A; Leschziner, Andres E

2015-01-01

The advent of a new generation of electron microscopes and direct electron detectors has realized the potential of single particle cryo-electron microscopy (cryo-EM) as a technique to generate high-resolution structures. Calculating these structures requires high performance computing clusters, a resource that may be limiting to many likely cryo-EM users. To address this limitation and facilitate the spread of cryo-EM, we developed a publicly available ‘off-the-shelf’ computing environment on Amazon's elastic cloud computing infrastructure. This environment provides users with single particle cryo-EM software packages and the ability to create computing clusters with 16–480+ CPUs. We tested our computing environment using a publicly available 80S yeast ribosome dataset and estimate that laboratories could determine high-resolution cryo-EM structures for $50 to $1500 per structure within a timeframe comparable to local clusters. Our analysis shows that Amazon's cloud computing environment may offer a viable computing environment for cryo-EM. DOI: http://dx.doi.org/10.7554/eLife.06664.001 PMID:25955969

A highly efficient multi-core algorithm for clustering extremely large datasets

PubMed Central

2010-01-01

Background In recent years, the demand for computational power in computational biology has increased due to rapidly growing data sets from microarray and other high-throughput technologies. This demand is likely to increase. Standard algorithms for analyzing data, such as cluster algorithms, need to be parallelized for fast processing. Unfortunately, most approaches for parallelizing algorithms largely rely on network communication protocols connecting and requiring multiple computers. One answer to this problem is to utilize the intrinsic capabilities in current multi-core hardware to distribute the tasks among the different cores of one computer. Results We introduce a multi-core parallelization of the k-means and k-modes cluster algorithms based on the design principles of transactional memory for clustering gene expression microarray type data and categorial SNP data. Our new shared memory parallel algorithms show to be highly efficient. We demonstrate their computational power and show their utility in cluster stability and sensitivity analysis employing repeated runs with slightly changed parameters. Computation speed of our Java based algorithm was increased by a factor of 10 for large data sets while preserving computational accuracy compared to single-core implementations and a recently published network based parallelization. Conclusions Most desktop computers and even notebooks provide at least dual-core processors. Our multi-core algorithms show that using modern algorithmic concepts, parallelization makes it possible to perform even such laborious tasks as cluster sensitivity and cluster number estimation on the laboratory computer. PMID:20370922

Unsupervised feature relevance analysis applied to improve ECG heartbeat clustering.

PubMed

Rodríguez-Sotelo, J L; Peluffo-Ordoñez, D; Cuesta-Frau, D; Castellanos-Domínguez, G

2012-10-01

The computer-assisted analysis of biomedical records has become an essential tool in clinical settings. However, current devices provide a growing amount of data that often exceeds the processing capacity of normal computers. As this amount of information rises, new demands for more efficient data extracting methods appear. This paper addresses the task of data mining in physiological records using a feature selection scheme. An unsupervised method based on relevance analysis is described. This scheme uses a least-squares optimization of the input feature matrix in a single iteration. The output of the algorithm is a feature weighting vector. The performance of the method was assessed using a heartbeat clustering test on real ECG records. The quantitative cluster validity measures yielded a correctly classified heartbeat rate of 98.69% (specificity), 85.88% (sensitivity) and 95.04% (general clustering performance), which is even higher than the performance achieved by other similar ECG clustering studies. The number of features was reduced on average from 100 to 18, and the temporal cost was a 43% lower than in previous ECG clustering schemes. Copyright © 2012 Elsevier Ireland Ltd. All rights reserved.

Enabling Diverse Software Stacks on Supercomputers using High Performance Virtual Clusters.

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

Younge, Andrew J.; Pedretti, Kevin; Grant, Ryan

While large-scale simulations have been the hallmark of the High Performance Computing (HPC) community for decades, Large Scale Data Analytics (LSDA) workloads are gaining attention within the scientific community not only as a processing component to large HPC simulations, but also as standalone scientific tools for knowledge discovery. With the path towards Exascale, new HPC runtime systems are also emerging in a way that differs from classical distributed com- puting models. However, system software for such capabilities on the latest extreme-scale DOE supercomputing needs to be enhanced to more appropriately support these types of emerging soft- ware ecosystems. In thismore » paper, we propose the use of Virtual Clusters on advanced supercomputing resources to enable systems to support not only HPC workloads, but also emerging big data stacks. Specifi- cally, we have deployed the KVM hypervisor within Cray's Compute Node Linux on a XC-series supercomputer testbed. We also use libvirt and QEMU to manage and provision VMs directly on compute nodes, leveraging Ethernet-over-Aries network emulation. To our knowledge, this is the first known use of KVM on a true MPP supercomputer. We investigate the overhead our solution using HPC benchmarks, both evaluating single-node performance as well as weak scaling of a 32-node virtual cluster. Overall, we find single node performance of our solution using KVM on a Cray is very efficient with near-native performance. However overhead increases by up to 20% as virtual cluster size increases, due to limitations of the Ethernet-over-Aries bridged network. Furthermore, we deploy Apache Spark with large data analysis workloads in a Virtual Cluster, ef- fectively demonstrating how diverse software ecosystems can be supported by High Performance Virtual Clusters.« less

Construction and Utilization of a Beowulf Computing Cluster: A User's Perspective

NASA Technical Reports Server (NTRS)

Woods, Judy L.; West, Jeff S.; Sulyma, Peter R.

2000-01-01

Lockheed Martin Space Operations - Stennis Programs (LMSO) at the John C Stennis Space Center (NASA/SSC) has designed and built a Beowulf computer cluster which is owned by NASA/SSC and operated by LMSO. The design and construction of the cluster are detailed in this paper. The cluster is currently used for Computational Fluid Dynamics (CFD) simulations. The CFD codes in use and their applications are discussed. Examples of some of the work are also presented. Performance benchmark studies have been conducted for the CFD codes being run on the cluster. The results of two of the studies are presented and discussed. The cluster is not currently being utilized to its full potential; therefore, plans are underway to add more capabilities. These include the addition of structural, thermal, fluid, and acoustic Finite Element Analysis codes as well as real-time data acquisition and processing during test operations at NASA/SSC. These plans are discussed as well.

Galaxy CloudMan: delivering cloud compute clusters.

PubMed

Afgan, Enis; Baker, Dannon; Coraor, Nate; Chapman, Brad; Nekrutenko, Anton; Taylor, James

2010-12-21

Widespread adoption of high-throughput sequencing has greatly increased the scale and sophistication of computational infrastructure needed to perform genomic research. An alternative to building and maintaining local infrastructure is "cloud computing", which, in principle, offers on demand access to flexible computational infrastructure. However, cloud computing resources are not yet suitable for immediate "as is" use by experimental biologists. We present a cloud resource management system that makes it possible for individual researchers to compose and control an arbitrarily sized compute cluster on Amazon's EC2 cloud infrastructure without any informatics requirements. Within this system, an entire suite of biological tools packaged by the NERC Bio-Linux team (http://nebc.nerc.ac.uk/tools/bio-linux) is available for immediate consumption. The provided solution makes it possible, using only a web browser, to create a completely configured compute cluster ready to perform analysis in less than five minutes. Moreover, we provide an automated method for building custom deployments of cloud resources. This approach promotes reproducibility of results and, if desired, allows individuals and labs to add or customize an otherwise available cloud system to better meet their needs. The expected knowledge and associated effort with deploying a compute cluster in the Amazon EC2 cloud is not trivial. The solution presented in this paper eliminates these barriers, making it possible for researchers to deploy exactly the amount of computing power they need, combined with a wealth of existing analysis software, to handle the ongoing data deluge.

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.

Collaborative Simulation Grid: Multiscale Quantum-Mechanical/Classical Atomistic Simulations on Distributed PC Clusters in the US and Japan

NASA Technical Reports Server (NTRS)

Kikuchi, Hideaki; Kalia, Rajiv; Nakano, Aiichiro; Vashishta, Priya; Iyetomi, Hiroshi; Ogata, Shuji; Kouno, Takahisa; Shimojo, Fuyuki; Tsuruta, Kanji; Saini, Subhash;

SAIL: Summation-bAsed Incremental Learning for Information-Theoretic Text Clustering.

PubMed

Cao, Jie; Wu, Zhiang; Wu, Junjie; Xiong, Hui

2013-04-01

Information-theoretic clustering aims to exploit information-theoretic measures as the clustering criteria. A common practice on this topic is the so-called Info-Kmeans, which performs K-means clustering with KL-divergence as the proximity function. While expert efforts on Info-Kmeans have shown promising results, a remaining challenge is to deal with high-dimensional sparse data such as text corpora. Indeed, it is possible that the centroids contain many zero-value features for high-dimensional text vectors, which leads to infinite KL-divergence values and creates a dilemma in assigning objects to centroids during the iteration process of Info-Kmeans. To meet this challenge, in this paper, we propose a Summation-bAsed Incremental Learning (SAIL) algorithm for Info-Kmeans clustering. Specifically, by using an equivalent objective function, SAIL replaces the computation of KL-divergence by the incremental computation of Shannon entropy. This can avoid the zero-feature dilemma caused by the use of KL-divergence. To improve the clustering quality, we further introduce the variable neighborhood search scheme and propose the V-SAIL algorithm, which is then accelerated by a multithreaded scheme in PV-SAIL. Our experimental results on various real-world text collections have shown that, with SAIL as a booster, the clustering performance of Info-Kmeans can be significantly improved. Also, V-SAIL and PV-SAIL indeed help improve the clustering quality at a lower cost of computation.

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

Application of the Linux cluster for exhaustive window haplotype analysis using the FBAT and Unphased programs.

PubMed

Mishima, Hiroyuki; Lidral, Andrew C; Ni, Jun

2008-05-28

Genetic association studies have been used to map disease-causing genes. A newly introduced statistical method, called exhaustive haplotype association study, analyzes genetic information consisting of different numbers and combinations of DNA sequence variations along a chromosome. Such studies involve a large number of statistical calculations and subsequently high computing power. It is possible to develop parallel algorithms and codes to perform the calculations on a high performance computing (HPC) system. However, most existing commonly-used statistic packages for genetic studies are non-parallel versions. Alternatively, one may use the cutting-edge technology of grid computing and its packages to conduct non-parallel genetic statistical packages on a centralized HPC system or distributed computing systems. In this paper, we report the utilization of a queuing scheduler built on the Grid Engine and run on a Rocks Linux cluster for our genetic statistical studies. Analysis of both consecutive and combinational window haplotypes was conducted by the FBAT (Laird et al., 2000) and Unphased (Dudbridge, 2003) programs. The dataset consisted of 26 loci from 277 extended families (1484 persons). Using the Rocks Linux cluster with 22 compute-nodes, FBAT jobs performed about 14.4-15.9 times faster, while Unphased jobs performed 1.1-18.6 times faster compared to the accumulated computation duration. Execution of exhaustive haplotype analysis using non-parallel software packages on a Linux-based system is an effective and efficient approach in terms of cost and performance.

Application of the Linux cluster for exhaustive window haplotype analysis using the FBAT and Unphased programs

PubMed Central

Mishima, Hiroyuki; Lidral, Andrew C; Ni, Jun

2008-01-01

Background Genetic association studies have been used to map disease-causing genes. A newly introduced statistical method, called exhaustive haplotype association study, analyzes genetic information consisting of different numbers and combinations of DNA sequence variations along a chromosome. Such studies involve a large number of statistical calculations and subsequently high computing power. It is possible to develop parallel algorithms and codes to perform the calculations on a high performance computing (HPC) system. However, most existing commonly-used statistic packages for genetic studies are non-parallel versions. Alternatively, one may use the cutting-edge technology of grid computing and its packages to conduct non-parallel genetic statistical packages on a centralized HPC system or distributed computing systems. In this paper, we report the utilization of a queuing scheduler built on the Grid Engine and run on a Rocks Linux cluster for our genetic statistical studies. Results Analysis of both consecutive and combinational window haplotypes was conducted by the FBAT (Laird et al., 2000) and Unphased (Dudbridge, 2003) programs. The dataset consisted of 26 loci from 277 extended families (1484 persons). Using the Rocks Linux cluster with 22 compute-nodes, FBAT jobs performed about 14.4–15.9 times faster, while Unphased jobs performed 1.1–18.6 times faster compared to the accumulated computation duration. Conclusion Execution of exhaustive haplotype analysis using non-parallel software packages on a Linux-based system is an effective and efficient approach in terms of cost and performance. PMID:18541045

Issues in ATM Support of High-Performance, Geographically Distributed Computing

NASA Technical Reports Server (NTRS)

Claus, Russell W.; Dowd, Patrick W.; Srinidhi, Saragur M.; Blade, Eric D.G

1995-01-01

This report experimentally assesses the effect of the underlying network in a cluster-based computing environment. The assessment is quantified by application-level benchmarking, process-level communication, and network file input/output. Two testbeds were considered, one small cluster of Sun workstations and another large cluster composed of 32 high-end IBM RS/6000 platforms. The clusters had Ethernet, fiber distributed data interface (FDDI), Fibre Channel, and asynchronous transfer mode (ATM) network interface cards installed, providing the same processors and operating system for the entire suite of experiments. The primary goal of this report is to assess the suitability of an ATM-based, local-area network to support interprocess communication and remote file input/output systems for distributed computing.

Dense, Efficient Chip-to-Chip Communication at the Extremes of Computing

ERIC Educational Resources Information Center

Loh, Matthew

2013-01-01

The scalability of CMOS technology has driven computation into a diverse range of applications across the power consumption, performance and size spectra. Communication is a necessary adjunct to computation, and whether this is to push data from node-to-node in a high-performance computing cluster or from the receiver of wireless link to a neural…

Grid Computing Environment using a Beowulf Cluster

NASA Astrophysics Data System (ADS)

Alanis, Fransisco; Mahmood, Akhtar

2003-10-01

Custom-made Beowulf clusters using PCs are currently replacing expensive supercomputers to carry out complex scientific computations. At the University of Texas - Pan American, we built a 8 Gflops Beowulf Cluster for doing HEP research using RedHat Linux 7.3 and the LAM-MPI middleware. We will describe how we built and configured our Cluster, which we have named the Sphinx Beowulf Cluster. We will describe the results of our cluster benchmark studies and the run-time plots of several parallel application codes that were compiled in C on the cluster using the LAM-XMPI graphics user environment. We will demonstrate a "simple" prototype grid environment, where we will submit and run parallel jobs remotely across multiple cluster nodes over the internet from the presentation room at Texas Tech. University. The Sphinx Beowulf Cluster will be used for monte-carlo grid test-bed studies for the LHC-ATLAS high energy physics experiment. Grid is a new IT concept for the next generation of the "Super Internet" for high-performance computing. The Grid will allow scientist worldwide to view and analyze huge amounts of data flowing from the large-scale experiments in High Energy Physics. The Grid is expected to bring together geographically and organizationally dispersed computational resources, such as CPUs, storage systems, communication systems, and data sources.

A 3D-PIV System for Gas Turbine Applications

NASA Astrophysics Data System (ADS)

Acharya, Sumanta

2002-08-01

Funds were received in April 2001 under the Department of Defense DURIP program for construction of a 48 processor high performance computing cluster. This report details the hardware, which was purchased, and how it has been used to enable and enhance research activities directly supported by, and of interest to, the Air Force Office of Scientific Research and the Department of Defense. The report is divided into two major sections. The first section after the summary describes the computer cluster, its setup, and some cluster hardware, and presents highlights of those efforts since installation of the cluster.

Galaxy CloudMan: delivering cloud compute clusters

PubMed Central

2010-01-01

Background Widespread adoption of high-throughput sequencing has greatly increased the scale and sophistication of computational infrastructure needed to perform genomic research. An alternative to building and maintaining local infrastructure is “cloud computing”, which, in principle, offers on demand access to flexible computational infrastructure. However, cloud computing resources are not yet suitable for immediate “as is” use by experimental biologists. Results We present a cloud resource management system that makes it possible for individual researchers to compose and control an arbitrarily sized compute cluster on Amazon’s EC2 cloud infrastructure without any informatics requirements. Within this system, an entire suite of biological tools packaged by the NERC Bio-Linux team (http://nebc.nerc.ac.uk/tools/bio-linux) is available for immediate consumption. The provided solution makes it possible, using only a web browser, to create a completely configured compute cluster ready to perform analysis in less than five minutes. Moreover, we provide an automated method for building custom deployments of cloud resources. This approach promotes reproducibility of results and, if desired, allows individuals and labs to add or customize an otherwise available cloud system to better meet their needs. Conclusions The expected knowledge and associated effort with deploying a compute cluster in the Amazon EC2 cloud is not trivial. The solution presented in this paper eliminates these barriers, making it possible for researchers to deploy exactly the amount of computing power they need, combined with a wealth of existing analysis software, to handle the ongoing data deluge. PMID:21210983

A fuzzy clustering algorithm to detect planar and quadric shapes

NASA Technical Reports Server (NTRS)

Krishnapuram, Raghu; Frigui, Hichem; Nasraoui, Olfa

1992-01-01

In this paper, we introduce a new fuzzy clustering algorithm to detect an unknown number of planar and quadric shapes in noisy data. The proposed algorithm is computationally and implementationally simple, and it overcomes many of the drawbacks of the existing algorithms that have been proposed for similar tasks. Since the clustering is performed in the original image space, and since no features need to be computed, this approach is particularly suited for sparse data. The algorithm may also be used in pattern recognition applications.

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.

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.

Visualization of unsteady computational fluid dynamics

NASA Astrophysics Data System (ADS)

Haimes, Robert

1994-11-01

A brief summary of the computer environment used for calculating three dimensional unsteady Computational Fluid Dynamic (CFD) results is presented. This environment requires a super computer as well as massively parallel processors (MPP's) and clusters of workstations acting as a single MPP (by concurrently working on the same task) provide the required computational bandwidth for CFD calculations of transient problems. The cluster of reduced instruction set computers (RISC) is a recent advent based on the low cost and high performance that workstation vendors provide. The cluster, with the proper software can act as a multiple instruction/multiple data (MIMD) machine. A new set of software tools is being designed specifically to address visualizing 3D unsteady CFD results in these environments. Three user's manuals for the parallel version of Visual3, pV3, revision 1.00 make up the bulk of this report.

Visualization of unsteady computational fluid dynamics

NASA Technical Reports Server (NTRS)

Haimes, Robert

1994-01-01

A brief summary of the computer environment used for calculating three dimensional unsteady Computational Fluid Dynamic (CFD) results is presented. This environment requires a super computer as well as massively parallel processors (MPP's) and clusters of workstations acting as a single MPP (by concurrently working on the same task) provide the required computational bandwidth for CFD calculations of transient problems. The cluster of reduced instruction set computers (RISC) is a recent advent based on the low cost and high performance that workstation vendors provide. The cluster, with the proper software can act as a multiple instruction/multiple data (MIMD) machine. A new set of software tools is being designed specifically to address visualizing 3D unsteady CFD results in these environments. Three user's manuals for the parallel version of Visual3, pV3, revision 1.00 make up the bulk of this report.

High-performance computing — an overview

NASA Astrophysics Data System (ADS)

Marksteiner, Peter

1996-08-01

An overview of high-performance computing (HPC) is given. Different types of computer architectures used in HPC are discussed: vector supercomputers, high-performance RISC processors, various parallel computers like symmetric multiprocessors, workstation clusters, massively parallel processors. Software tools and programming techniques used in HPC are reviewed: vectorizing compilers, optimization and vector tuning, optimization for RISC processors; parallel programming techniques like shared-memory parallelism, message passing and data parallelism; and numerical libraries.

A Parallel Processing Algorithm for Remote Sensing Classification

NASA Technical Reports Server (NTRS)

Gualtieri, J. Anthony

2005-01-01

A current thread in parallel computation is the use of cluster computers created by networking a few to thousands of commodity general-purpose workstation-level commuters using the Linux operating system. For example on the Medusa cluster at NASA/GSFC, this provides for super computing performance, 130 G(sub flops) (Linpack Benchmark) at moderate cost, $370K. However, to be useful for scientific computing in the area of Earth science, issues of ease of programming, access to existing scientific libraries, and portability of existing code need to be considered. In this paper, I address these issues in the context of tools for rendering earth science remote sensing data into useful products. In particular, I focus on a problem that can be decomposed into a set of independent tasks, which on a serial computer would be performed sequentially, but with a cluster computer can be performed in parallel, giving an obvious speedup. To make the ideas concrete, I consider the problem of classifying hyperspectral imagery where some ground truth is available to train the classifier. In particular I will use the Support Vector Machine (SVM) approach as applied to hyperspectral imagery. The approach will be to introduce notions about parallel computation and then to restrict the development to the SVM problem. Pseudocode (an outline of the computation) will be described and then details specific to the implementation will be given. Then timing results will be reported to show what speedups are possible using parallel computation. The paper will close with a discussion of the results.

A parallel-processing approach to computing for the geographic sciences

USGS Publications Warehouse

Crane, Michael; Steinwand, Dan; Beckmann, Tim; Krpan, Greg; Haga, Jim; Maddox, Brian; Feller, Mark

2001-01-01

The overarching goal of this project is to build a spatially distributed infrastructure for information science research by forming a team of information science researchers and providing them with similar hardware and software tools to perform collaborative research. Four geographically distributed Centers of the U.S. Geological Survey (USGS) are developing their own clusters of low-cost personal computers into parallel computing environments that provide a costeffective way for the USGS to increase participation in the high-performance computing community. Referred to as Beowulf clusters, these hybrid systems provide the robust computing power required for conducting research into various areas, such as advanced computer architecture, algorithms to meet the processing needs for real-time image and data processing, the creation of custom datasets from seamless source data, rapid turn-around of products for emergency response, and support for computationally intense spatial and temporal modeling.

Method and apparatus for offloading compute resources to a flash co-processing appliance

DOEpatents

Tzelnic, Percy; Faibish, Sorin; Gupta, Uday K.; Bent, John; Grider, Gary Alan; Chen, Hsing -bung

2015-10-13

Solid-State Drive (SSD) burst buffer nodes are interposed into a parallel supercomputing cluster to enable fast burst checkpoint of cluster memory to or from nearby interconnected solid-state storage with asynchronous migration between the burst buffer nodes and slower more distant disk storage. The SSD nodes also perform tasks offloaded from the compute nodes or associated with the checkpoint data. For example, the data for the next job is preloaded in the SSD node and very fast uploaded to the respective compute node just before the next job starts. During a job, the SSD nodes perform fast visualization and statistical analysis upon the checkpoint data. The SSD nodes can also perform data reduction and encryption of the checkpoint data.

Using Agent Base Models to Optimize Large Scale Network for Large System Inventories

NASA Technical Reports Server (NTRS)

Shameldin, Ramez Ahmed; Bowling, Shannon R.

2010-01-01

The aim of this paper is to use Agent Base Models (ABM) to optimize large scale network handling capabilities for large system inventories and to implement strategies for the purpose of reducing capital expenses. The models used in this paper either use computational algorithms or procedure implementations developed by Matlab to simulate agent based models in a principal programming language and mathematical theory using clusters, these clusters work as a high performance computational performance to run the program in parallel computational. In both cases, a model is defined as compilation of a set of structures and processes assumed to underlie the behavior of a network system.

GREEN SUPERCOMPUTING IN A DESKTOP BOX

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

HSU, CHUNG-HSING; FENG, WU-CHUN; CHING, AVERY

2007-01-17

The computer workstation, introduced by Sun Microsystems in 1982, was the tool of choice for scientists and engineers as an interactive computing environment for the development of scientific codes. However, by the mid-1990s, the performance of workstations began to lag behind high-end commodity PCs. This, coupled with the disappearance of BSD-based operating systems in workstations and the emergence of Linux as an open-source operating system for PCs, arguably led to the demise of the workstation as we knew it. Around the same time, computational scientists started to leverage PCs running Linux to create a commodity-based (Beowulf) cluster that provided dedicatedmore » computer cycles, i.e., supercomputing for the rest of us, as a cost-effective alternative to large supercomputers, i.e., supercomputing for the few. However, as the cluster movement has matured, with respect to cluster hardware and open-source software, these clusters have become much more like their large-scale supercomputing brethren - a shared (and power-hungry) datacenter resource that must reside in a machine-cooled room in order to operate properly. Consequently, the above observations, when coupled with the ever-increasing performance gap between the PC and cluster supercomputer, provide the motivation for a 'green' desktop supercomputer - a turnkey solution that provides an interactive and parallel computing environment with the approximate form factor of a Sun SPARCstation 1 'pizza box' workstation. In this paper, they present the hardware and software architecture of such a solution as well as its prowess as a developmental platform for parallel codes. In short, imagine a 12-node personal desktop supercomputer that achieves 14 Gflops on Linpack but sips only 185 watts of power at load, resulting in a performance-power ratio that is over 300% better than their reference SMP platform.« less

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.

Systematic exploration of unsupervised methods for mapping behavior

NASA Astrophysics Data System (ADS)

Todd, Jeremy G.; Kain, Jamey S.; de Bivort, Benjamin L.

2017-02-01

To fully understand the mechanisms giving rise to behavior, we need to be able to precisely measure it. When coupled with large behavioral data sets, unsupervised clustering methods offer the potential of unbiased mapping of behavioral spaces. However, unsupervised techniques to map behavioral spaces are in their infancy, and there have been few systematic considerations of all the methodological options. We compared the performance of seven distinct mapping methods in clustering a wavelet-transformed data set consisting of the x- and y-positions of the six legs of individual flies. Legs were automatically tracked by small pieces of fluorescent dye, while the fly was tethered and walking on an air-suspended ball. We find that there is considerable variation in the performance of these mapping methods, and that better performance is attained when clustering is done in higher dimensional spaces (which are otherwise less preferable because they are hard to visualize). High dimensionality means that some algorithms, including the non-parametric watershed cluster assignment algorithm, cannot be used. We developed an alternative watershed algorithm which can be used in high-dimensional spaces when a probability density estimate can be computed directly. With these tools in hand, we examined the behavioral space of fly leg postural dynamics and locomotion. We find a striking division of behavior into modes involving the fore legs and modes involving the hind legs, with few direct transitions between them. By computing behavioral clusters using the data from all flies simultaneously, we show that this division appears to be common to all flies. We also identify individual-to-individual differences in behavior and behavioral transitions. Lastly, we suggest a computational pipeline that can achieve satisfactory levels of performance without the taxing computational demands of a systematic combinatorial approach.

Performance enhancement of a web-based picture archiving and communication system using commercial off-the-shelf server clusters.

PubMed

Liu, Yan-Lin; Shih, Cheng-Ting; Chang, Yuan-Jen; Chang, Shu-Jun; Wu, Jay

2014-01-01

The rapid development of picture archiving and communication systems (PACSs) thoroughly changes the way of medical informatics communication and management. However, as the scale of a hospital's operations increases, the large amount of digital images transferred in the network inevitably decreases system efficiency. In this study, a server cluster consisting of two server nodes was constructed. Network load balancing (NLB), distributed file system (DFS), and structured query language (SQL) duplication services were installed. A total of 1 to 16 workstations were used to transfer computed radiography (CR), computed tomography (CT), and magnetic resonance (MR) images simultaneously to simulate the clinical situation. The average transmission rate (ATR) was analyzed between the cluster and noncluster servers. In the download scenario, the ATRs of CR, CT, and MR images increased by 44.3%, 56.6%, and 100.9%, respectively, when using the server cluster, whereas the ATRs increased by 23.0%, 39.2%, and 24.9% in the upload scenario. In the mix scenario, the transmission performance increased by 45.2% when using eight computer units. The fault tolerance mechanisms of the server cluster maintained the system availability and image integrity. The server cluster can improve the transmission efficiency while maintaining high reliability and continuous availability in a healthcare environment.

Massively parallel unsupervised single-particle cryo-EM data clustering via statistical manifold learning

PubMed Central

Wu, Jiayi; Ma, Yong-Bei; Congdon, Charles; Brett, Bevin; Chen, Shuobing; Xu, Yaofang; Ouyang, Qi

2017-01-01

Structural heterogeneity in single-particle cryo-electron microscopy (cryo-EM) data represents a major challenge for high-resolution structure determination. Unsupervised classification may serve as the first step in the assessment of structural heterogeneity. However, traditional algorithms for unsupervised classification, such as K-means clustering and maximum likelihood optimization, may classify images into wrong classes with decreasing signal-to-noise-ratio (SNR) in the image data, yet demand increased computational costs. Overcoming these limitations requires further development of clustering algorithms for high-performance cryo-EM data processing. Here we introduce an unsupervised single-particle clustering algorithm derived from a statistical manifold learning framework called generative topographic mapping (GTM). We show that unsupervised GTM clustering improves classification accuracy by about 40% in the absence of input references for data with lower SNRs. Applications to several experimental datasets suggest that our algorithm can detect subtle structural differences among classes via a hierarchical clustering strategy. After code optimization over a high-performance computing (HPC) environment, our software implementation was able to generate thousands of reference-free class averages within hours in a massively parallel fashion, which allows a significant improvement on ab initio 3D reconstruction and assists in the computational purification of homogeneous datasets for high-resolution visualization. PMID:28786986

Massively parallel unsupervised single-particle cryo-EM data clustering via statistical manifold learning.

PubMed

Wu, Jiayi; Ma, Yong-Bei; Congdon, Charles; Brett, Bevin; Chen, Shuobing; Xu, Yaofang; Ouyang, Qi; Mao, Youdong

2017-01-01

Structural heterogeneity in single-particle cryo-electron microscopy (cryo-EM) data represents a major challenge for high-resolution structure determination. Unsupervised classification may serve as the first step in the assessment of structural heterogeneity. However, traditional algorithms for unsupervised classification, such as K-means clustering and maximum likelihood optimization, may classify images into wrong classes with decreasing signal-to-noise-ratio (SNR) in the image data, yet demand increased computational costs. Overcoming these limitations requires further development of clustering algorithms for high-performance cryo-EM data processing. Here we introduce an unsupervised single-particle clustering algorithm derived from a statistical manifold learning framework called generative topographic mapping (GTM). We show that unsupervised GTM clustering improves classification accuracy by about 40% in the absence of input references for data with lower SNRs. Applications to several experimental datasets suggest that our algorithm can detect subtle structural differences among classes via a hierarchical clustering strategy. After code optimization over a high-performance computing (HPC) environment, our software implementation was able to generate thousands of reference-free class averages within hours in a massively parallel fashion, which allows a significant improvement on ab initio 3D reconstruction and assists in the computational purification of homogeneous datasets for high-resolution visualization.

Performance Enhancement of a Web-Based Picture Archiving and Communication System Using Commercial Off-the-Shelf Server Clusters

PubMed Central

Chang, Shu-Jun; Wu, Jay

2014-01-01

The rapid development of picture archiving and communication systems (PACSs) thoroughly changes the way of medical informatics communication and management. However, as the scale of a hospital's operations increases, the large amount of digital images transferred in the network inevitably decreases system efficiency. In this study, a server cluster consisting of two server nodes was constructed. Network load balancing (NLB), distributed file system (DFS), and structured query language (SQL) duplication services were installed. A total of 1 to 16 workstations were used to transfer computed radiography (CR), computed tomography (CT), and magnetic resonance (MR) images simultaneously to simulate the clinical situation. The average transmission rate (ATR) was analyzed between the cluster and noncluster servers. In the download scenario, the ATRs of CR, CT, and MR images increased by 44.3%, 56.6%, and 100.9%, respectively, when using the server cluster, whereas the ATRs increased by 23.0%, 39.2%, and 24.9% in the upload scenario. In the mix scenario, the transmission performance increased by 45.2% when using eight computer units. The fault tolerance mechanisms of the server cluster maintained the system availability and image integrity. The server cluster can improve the transmission efficiency while maintaining high reliability and continuous availability in a healthcare environment. PMID:24701580

Optimising the Parallelisation of OpenFOAM Simulations

DTIC Science & Technology

2014-06-01

UNCLASSIFIED UNCLASSIFIED Optimising the Parallelisation of OpenFOAM Simulations Shannon Keough Maritime Division Defence...Science and Technology Organisation DSTO-TR-2987 ABSTRACT The OpenFOAM computational fluid dynamics toolbox allows parallel computation of...performance of a given high performance computing cluster with several OpenFOAM cases, running using a combination of MPI libraries and corresponding MPI

Multi-hop routing mechanism for reliable sensor computing.

PubMed

Chen, Jiann-Liang; Ma, Yi-Wei; Lai, Chia-Ping; Hu, Chia-Cheng; Huang, Yueh-Min

2009-01-01

Current research on routing in wireless sensor computing concentrates on increasing the service lifetime, enabling scalability for large number of sensors and supporting fault tolerance for battery exhaustion and broken nodes. A sensor node is naturally exposed to various sources of unreliable communication channels and node failures. Sensor nodes have many failure modes, and each failure degrades the network performance. This work develops a novel mechanism, called Reliable Routing Mechanism (RRM), based on a hybrid cluster-based routing protocol to specify the best reliable routing path for sensor computing. Table-driven intra-cluster routing and on-demand inter-cluster routing are combined by changing the relationship between clusters for sensor computing. Applying a reliable routing mechanism in sensor computing can improve routing reliability, maintain low packet loss, minimize management overhead and save energy consumption. Simulation results indicate that the reliability of the proposed RRM mechanism is around 25% higher than that of the Dynamic Source Routing (DSR) and ad hoc On-demand Distance Vector routing (AODV) mechanisms.

Redirecting Under-Utilised Computer Laboratories into Cluster Computing Facilities

ERIC Educational Resources Information Center

Atkinson, John S.; Spenneman, Dirk H. R.; Cornforth, David

2005-01-01

Purpose: To provide administrators at an Australian university with data on the feasibility of redirecting under-utilised computer laboratories facilities into a distributed high performance computing facility. Design/methodology/approach: The individual log-in records for each computer located in the computer laboratories at the university were…

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 .

Jade: using on-demand cloud analysis to give scientists back their flow

NASA Astrophysics Data System (ADS)

Robinson, N.; Tomlinson, J.; Hilson, A. J.; Arribas, A.; Powell, T.

2017-12-01

The UK's Met Office generates 400 TB weather and climate data every day by running physical models on its Top 20 supercomputer. As data volumes explode, there is a danger that analysis workflows become dominated by watching progress bars, and not thinking about science. We have been researching how we can use distributed computing to allow analysts to process these large volumes of high velocity data in a way that's easy, effective and cheap.Our prototype analysis stack, Jade, tries to encapsulate this. Functionality includes: An under-the-hood Dask engine which parallelises and distributes computations, without the need to retrain analysts Hybrid compute clusters (AWS, Alibaba, and local compute) comprising many thousands of cores Clusters which autoscale up/down in response to calculation load using Kubernetes, and balances the cluster across providers based on the current price of compute Lazy data access from cloud storage via containerised OpenDAP This technology stack allows us to perform calculations many orders of magnitude faster than is possible on local workstations. It is also possible to outperform dedicated local compute clusters, as cloud compute can, in principle, scale to much larger scales. The use of ephemeral compute resources also makes this implementation cost efficient.

Gate sequence for continuous variable one-way quantum computation

PubMed Central

Su, Xiaolong; Hao, Shuhong; Deng, Xiaowei; Ma, Lingyu; Wang, Meihong; Jia, Xiaojun; Xie, Changde; Peng, Kunchi

2013-01-01

Measurement-based one-way quantum computation using cluster states as resources provides an efficient model to perform computation and information processing of quantum codes. Arbitrary Gaussian quantum computation can be implemented sufficiently by long single-mode and two-mode gate sequences. However, continuous variable gate sequences have not been realized so far due to an absence of cluster states larger than four submodes. Here we present the first continuous variable gate sequence consisting of a single-mode squeezing gate and a two-mode controlled-phase gate based on a six-mode cluster state. The quantum property of this gate sequence is confirmed by the fidelities and the quantum entanglement of two output modes, which depend on both the squeezing and controlled-phase gates. The experiment demonstrates the feasibility of implementing Gaussian quantum computation by means of accessible gate sequences.

Assessing the Amazon Cloud Suitability for CLARREO's Computational Needs

NASA Technical Reports Server (NTRS)

Goldin, Daniel; Vakhnin, Andrei A.; Currey, Jon C.

2015-01-01

In this document we compare the performance of the Amazon Web Services (AWS), also known as Amazon Cloud, with the CLARREO (Climate Absolute Radiance and Refractivity Observatory) cluster and assess its suitability for computational needs of the CLARREO mission. A benchmark executable to process one month and one year of PARASOL (Polarization and Anistropy of Reflectances for Atmospheric Sciences coupled with Observations from a Lidar) data was used. With the optimal AWS configuration, adequate data-processing times, comparable to the CLARREO cluster, were found. The assessment of alternatives to the CLARREO cluster continues and several options, such as a NASA-based cluster, are being considered.

CLUSFAVOR 5.0: hierarchical cluster and principal-component analysis of microarray-based transcriptional profiles

PubMed Central

Peterson, Leif E

2002-01-01

CLUSFAVOR (CLUSter and Factor Analysis with Varimax Orthogonal Rotation) 5.0 is a Windows-based computer program for hierarchical cluster and principal-component analysis of microarray-based transcriptional profiles. CLUSFAVOR 5.0 standardizes input data; sorts data according to gene-specific coefficient of variation, standard deviation, average and total expression, and Shannon entropy; performs hierarchical cluster analysis using nearest-neighbor, unweighted pair-group method using arithmetic averages (UPGMA), or furthest-neighbor joining methods, and Euclidean, correlation, or jack-knife distances; and performs principal-component analysis. PMID:12184816

Scaling predictive modeling in drug development with cloud computing.

PubMed

Moghadam, Behrooz Torabi; Alvarsson, Jonathan; Holm, Marcus; Eklund, Martin; Carlsson, Lars; Spjuth, Ola

2015-01-26

Growing data sets with increased time for analysis is hampering predictive modeling in drug discovery. Model building can be carried out on high-performance computer clusters, but these can be expensive to purchase and maintain. We have evaluated ligand-based modeling on cloud computing resources where computations are parallelized and run on the Amazon Elastic Cloud. We trained models on open data sets of varying sizes for the end points logP and Ames mutagenicity and compare with model building parallelized on a traditional high-performance computing cluster. We show that while high-performance computing results in faster model building, the use of cloud computing resources is feasible for large data sets and scales well within cloud instances. An additional advantage of cloud computing is that the costs of predictive models can be easily quantified, and a choice can be made between speed and economy. The easy access to computational resources with no up-front investments makes cloud computing an attractive alternative for scientists, especially for those without access to a supercomputer, and our study shows that it enables cost-efficient modeling of large data sets on demand within reasonable time.

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.

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.

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.

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.

m-BIRCH: an online clustering approach for computer vision applications

NASA Astrophysics Data System (ADS)

Madan, Siddharth K.; Dana, Kristin J.

2015-03-01

We adapt a classic online clustering algorithm called Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH), to incrementally cluster large datasets of features commonly used in multimedia and computer vision. We call the adapted version modified-BIRCH (m-BIRCH). The algorithm uses only a fraction of the dataset memory to perform clustering, and updates the clustering decisions when new data comes in. Modifications made in m-BIRCH enable data driven parameter selection and effectively handle varying density regions in the feature space. Data driven parameter selection automatically controls the level of coarseness of the data summarization. Effective handling of varying density regions is necessary to well represent the different density regions in data summarization. We use m-BIRCH to cluster 840K color SIFT descriptors, and 60K outlier corrupted grayscale patches. We use the algorithm to cluster datasets consisting of challenging non-convex clustering patterns. Our implementation of the algorithm provides an useful clustering tool and is made publicly available.

Implementation of the force decomposition machine for molecular dynamics simulations.

PubMed

Borštnik, Urban; Miller, Benjamin T; Brooks, Bernard R; Janežič, Dušanka

2012-09-01

We present the design and implementation of the force decomposition machine (FDM), a cluster of personal computers (PCs) that is tailored to running molecular dynamics (MD) simulations using the distributed diagonal force decomposition (DDFD) parallelization method. The cluster interconnect architecture is optimized for the communication pattern of the DDFD method. Our implementation of the FDM relies on standard commodity components even for networking. Although the cluster is meant for DDFD MD simulations, it remains general enough for other parallel computations. An analysis of several MD simulation runs on both the FDM and a standard PC cluster demonstrates that the FDM's interconnect architecture provides a greater performance compared to a more general cluster interconnect. Copyright © 2012 Elsevier Inc. All rights reserved.

Data Intensive Computing on Amazon Web Services

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

Magana-Zook, S. A.

The Geophysical Monitoring Program (GMP) has spent the past few years building up the capability to perform data intensive computing using what have been referred to as “big data” tools. These big data tools would be used against massive archives of seismic signals (>300 TB) to conduct research not previously possible. Examples of such tools include Hadoop (HDFS, MapReduce), HBase, Hive, Storm, Spark, Solr, and many more by the day. These tools are useful for performing data analytics on datasets that exceed the resources of traditional analytic approaches. To this end, a research big data cluster (“Cluster A”) was setmore » up as a collaboration between GMP and Livermore Computing (LC).« less

Performance Evaluation in Network-Based Parallel Computing

NASA Technical Reports Server (NTRS)

Dezhgosha, Kamyar

1996-01-01

Network-based parallel computing is emerging as a cost-effective alternative for solving many problems which require use of supercomputers or massively parallel computers. The primary objective of this project has been to conduct experimental research on performance evaluation for clustered parallel computing. First, a testbed was established by augmenting our existing SUNSPARCs' network with PVM (Parallel Virtual Machine) which is a software system for linking clusters of machines. Second, a set of three basic applications were selected. The applications consist of a parallel search, a parallel sort, a parallel matrix multiplication. These application programs were implemented in C programming language under PVM. Third, we conducted performance evaluation under various configurations and problem sizes. Alternative parallel computing models and workload allocations for application programs were explored. The performance metric was limited to elapsed time or response time which in the context of parallel computing can be expressed in terms of speedup. The results reveal that the overhead of communication latency between processes in many cases is the restricting factor to performance. That is, coarse-grain parallelism which requires less frequent communication between processes will result in higher performance in network-based computing. Finally, we are in the final stages of installing an Asynchronous Transfer Mode (ATM) switch and four ATM interfaces (each 155 Mbps) which will allow us to extend our study to newer applications, performance metrics, and configurations.

Using Cluster Bootstrapping to Analyze Nested Data With a Few Clusters.

PubMed

Huang, Francis L

2018-04-01

Cluster randomized trials involving participants nested within intact treatment and control groups are commonly performed in various educational, psychological, and biomedical studies. However, recruiting and retaining intact groups present various practical, financial, and logistical challenges to evaluators and often, cluster randomized trials are performed with a low number of clusters (~20 groups). Although multilevel models are often used to analyze nested data, researchers may be concerned of potentially biased results due to having only a few groups under study. Cluster bootstrapping has been suggested as an alternative procedure when analyzing clustered data though it has seen very little use in educational and psychological studies. Using a Monte Carlo simulation that varied the number of clusters, average cluster size, and intraclass correlations, we compared standard errors using cluster bootstrapping with those derived using ordinary least squares regression and multilevel models. Results indicate that cluster bootstrapping, though more computationally demanding, can be used as an alternative procedure for the analysis of clustered data when treatment effects at the group level are of primary interest. Supplementary material showing how to perform cluster bootstrapped regressions using R is also provided.

High Performance Distributed Computing in a Supercomputer Environment: Computational Services and Applications Issues

NASA Technical Reports Server (NTRS)

Kramer, Williams T. C.; Simon, Horst D.

1994-01-01

This tutorial proposes to be a practical guide for the uninitiated to the main topics and themes of high-performance computing (HPC), with particular emphasis to distributed computing. The intent is first to provide some guidance and directions in the rapidly increasing field of scientific computing using both massively parallel and traditional supercomputers. Because of their considerable potential computational power, loosely or tightly coupled clusters of workstations are increasingly considered as a third alternative to both the more conventional supercomputers based on a small number of powerful vector processors, as well as high massively parallel processors. Even though many research issues concerning the effective use of workstation clusters and their integration into a large scale production facility are still unresolved, such clusters are already used for production computing. In this tutorial we will utilize the unique experience made at the NAS facility at NASA Ames Research Center. Over the last five years at NAS massively parallel supercomputers such as the Connection Machines CM-2 and CM-5 from Thinking Machines Corporation and the iPSC/860 (Touchstone Gamma Machine) and Paragon Machines from Intel were used in a production supercomputer center alongside with traditional vector supercomputers such as the Cray Y-MP and C90.

Computational intelligence for the Balanced Scorecard: studying performance trends of hemodialysis clinics.

PubMed

Cattinelli, Isabella; Bolzoni, Elena; Chermisi, Milena; Bellocchio, Francesco; Barbieri, Carlo; Mari, Flavio; Amato, Claudia; Menzer, Marcus; Stopper, Andrea; Gatti, Emanuele

2013-07-01

The Balanced Scorecard (BSC) is a general, widely employed instrument for enterprise performance monitoring based on the periodic assessment of strategic Key Performance Indicators that are scored against preset targets. The BSC is currently employed as an effective management support tool within Fresenius Medical Care (FME) and is routinely analyzed via standard statistical methods. More recently, the application of computational intelligence techniques (namely, self-organizing maps) to BSC data has been proposed as a way to enhance the quantity and quality of information that can be extracted from it. In this work, additional methods are presented to analyze the evolution of clinic performance over time. Performance evolution is studied at the single-clinic level by computing two complementary indexes that measure the proportion of time spent within performance clusters and improving/worsening trends. Self-organizing maps are used in conjunction with these indexes to identify the specific drivers of the observed performance. The performance evolution for groups of clinics is modeled under a probabilistic framework by resorting to Markov chain properties. These allow a study of the probability of transitioning between performance clusters as time progresses for the identification of the performance level that is expected to become dominant over time. We show the potential of the proposed methods through illustrative results derived from the analysis of BSC data of 109 FME clinics in three countries. We were able to identify the performance drivers for specific groups of clinics and to distinguish between countries whose performances are likely to improve from those where a decline in performance might be expected. According to the stationary distribution of the Markov chain, the expected trend is best in Turkey (where the highest performance cluster has the highest probability, P=0.46), followed by Portugal (where the second best performance cluster dominates, with P=0.50), and finally Italy (where the second best performance cluster has P=0.34). These results highlight the ability of the proposed methods to extract insights about performance trends that cannot be easily extrapolated using standard analyses and that are valuable in directing management strategies within a continuous quality improvement policy. Copyright © 2013 Elsevier B.V. All rights reserved.

Upgrading of the LGD cluster at JINR to support DLNP experiments

NASA Astrophysics Data System (ADS)

Bednyakov, I. V.; Dolbilov, A. G.; Ivanov, Yu. P.

2017-01-01

Since its construction in 2005, the Computing Cluster of the Dzhelepov Laboratory of Nuclear Problems has been mainly used to perform calculations (data analysis, simulation, etc.) for various scientific collaborations in which DLNP scientists take an active part. The Cluster also serves to train specialists. Much has changed in the past decades, and the necessity has arisen to upgrade the cluster, increasing its power and replacing the outdated equipment to maintain its reliability and modernity. In this work we describe the experience of performing this upgrading, which can be helpful for system administrators to put new equipment for clusters of this type into operation quickly and efficiently.

Monitoring by Use of Clusters of Sensor-Data Vectors

NASA Technical Reports Server (NTRS)

Iverson, David L.

2007-01-01

The inductive monitoring system (IMS) is a system of computer hardware and software for automated monitoring of the performance, operational condition, physical integrity, and other aspects of the health of a complex engineering system (e.g., an industrial process line or a spacecraft). The input to the IMS consists of streams of digitized readings from sensors in the monitored system. The IMS determines the type and amount of any deviation of the monitored system from a nominal or normal ( healthy ) condition on the basis of a comparison between (1) vectors constructed from the incoming sensor data and (2) corresponding vectors in a database of nominal or normal behavior. The term inductive reflects the use of a process reminiscent of traditional mathematical induction to learn about normal operation and build the nominal-condition database. The IMS offers two major advantages over prior computational monitoring systems: The computational burden of the IMS is significantly smaller, and there is no need for abnormal-condition sensor data for training the IMS to recognize abnormal conditions. The figure schematically depicts the relationships among the computational processes effected by the IMS. Training sensor data are gathered during normal operation of the monitored system, detailed computational simulation of operation of the monitored system, or both. The training data are formed into vectors that are used to generate the database. The vectors in the database are clustered into regions that represent normal or nominal operation. Once the database has been generated, the IMS compares the vectors of incoming sensor data with vectors representative of the clusters. The monitored system is deemed to be operating normally or abnormally, depending on whether the vector of incoming sensor data is or is not, respectively, sufficiently close to one of the clusters. For this purpose, a distance between two vectors is calculated by a suitable metric (e.g., Euclidean distance) and "sufficiently close" signifies lying at a distance less than a specified threshold value. It must be emphasized that although the IMS is intended to detect off-nominal or abnormal performance or health, it is not necessarily capable of performing a thorough or detailed diagnosis. Limited diagnostic information may be available under some circumstances. For example, the distance of a vector of incoming sensor data from the nearest cluster could serve as an indication of the severity of a malfunction. The identity of the nearest cluster may be a clue as to the identity of the malfunctioning component or subsystem. It is possible to decrease the IMS computation time by use of a combination of cluster-indexing and -retrieval methods. For example, in one method, the distances between each cluster and two or more reference vectors can be used for the purpose of indexing and retrieval. The clusters are sorted into a list according to these distance values, typically in ascending order of distance. When a set of input data arrives and is to be tested, the data are first arranged as an ordered set (that is, a vector). The distances from the input vector to the reference points are computed. The search of clusters from the list can then be limited to those clusters lying within a certain distance range from the input vector; the computation time is reduced by not searching the clusters at a greater distance.

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

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.

HipMCL: a high-performance parallel implementation of the Markov clustering algorithm for large-scale networks

PubMed Central

Azad, Ariful; Ouzounis, Christos A; Kyrpides, Nikos C; Buluç, Aydin

2018-01-01

Abstract Biological networks capture structural or functional properties of relevant entities such as molecules, proteins or genes. Characteristic examples are gene expression networks or protein–protein interaction networks, which hold information about functional affinities or structural similarities. Such networks have been expanding in size due to increasing scale and abundance of biological data. While various clustering algorithms have been proposed to find highly connected regions, Markov Clustering (MCL) has been one of the most successful approaches to cluster sequence similarity or expression networks. Despite its popularity, MCL’s scalability to cluster large datasets still remains a bottleneck due to high running times and memory demands. Here, we present High-performance MCL (HipMCL), a parallel implementation of the original MCL algorithm that can run on distributed-memory computers. We show that HipMCL can efficiently utilize 2000 compute nodes and cluster a network of ∼70 million nodes with ∼68 billion edges in ∼2.4 h. By exploiting distributed-memory environments, HipMCL clusters large-scale networks several orders of magnitude faster than MCL and enables clustering of even bigger networks. HipMCL is based on MPI and OpenMP and is freely available under a modified BSD license. PMID:29315405

Simulations of the Formation and Evolution of X-ray Clusters

NASA Astrophysics Data System (ADS)

Bryan, G. L.; Klypin, A.; Norman, M. L.

1994-05-01

We describe results from a set of Omega = 1 Cold plus Hot Dark Matter (CHDM) and Cold Dark Matter (CDM) simulations. We examine the formation and evolution of X-ray clusters in a cosmological setting with sufficient numbers to perform statistical analysis. We find that CDM, normalized to COBE, seems to produce too many large clusters, both in terms of the luminosity (dn/dL) and temperature (dn/dT) functions. The CHDM simulation produces fewer clusters and the temperature distribution (our numerically most secure result) matches observations where they overlap. The computed cluster luminosity function drops below observations, but we are almost surely underestimating the X-ray luminosity. Because of the lower fluctuations in CHDM, there are only a small number of bright clusters in our simulation volume; however we can use the simulated clusters to fix the relation between temperature and velocity dispersion, allowing us to use collisionless N-body codes to probe larger length scales with correspondingly brighter clusters. The hydrodynamic simulations have been performed with a hybrid particle-mesh scheme for the dark matter and a high resolution grid-based piecewise parabolic method for the adiabatic gas dynamics. This combination has been implemented for massively parallel computers, allowing us to achive grids as large as 512(3) .

HipMCL: a high-performance parallel implementation of the Markov clustering algorithm for large-scale networks

DOE PAGES

Azad, Ariful; Pavlopoulos, Georgios A.; Ouzounis, Christos A.; ...

2018-01-05

Biological networks capture structural or functional properties of relevant entities such as molecules, proteins or genes. Characteristic examples are gene expression networks or protein–protein interaction networks, which hold information about functional affinities or structural similarities. Such networks have been expanding in size due to increasing scale and abundance of biological data. While various clustering algorithms have been proposed to find highly connected regions, Markov Clustering (MCL) has been one of the most successful approaches to cluster sequence similarity or expression networks. Despite its popularity, MCL’s scalability to cluster large datasets still remains a bottleneck due to high running times andmore » memory demands. In this paper, we present High-performance MCL (HipMCL), a parallel implementation of the original MCL algorithm that can run on distributed-memory computers. We show that HipMCL can efficiently utilize 2000 compute nodes and cluster a network of ~70 million nodes with ~68 billion edges in ~2.4 h. By exploiting distributed-memory environments, HipMCL clusters large-scale networks several orders of magnitude faster than MCL and enables clustering of even bigger networks. Finally, HipMCL is based on MPI and OpenMP and is freely available under a modified BSD license.« less

HipMCL: a high-performance parallel implementation of the Markov clustering algorithm for large-scale networks

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

Azad, Ariful; Pavlopoulos, Georgios A.; Ouzounis, Christos A.

Biological networks capture structural or functional properties of relevant entities such as molecules, proteins or genes. Characteristic examples are gene expression networks or protein–protein interaction networks, which hold information about functional affinities or structural similarities. Such networks have been expanding in size due to increasing scale and abundance of biological data. While various clustering algorithms have been proposed to find highly connected regions, Markov Clustering (MCL) has been one of the most successful approaches to cluster sequence similarity or expression networks. Despite its popularity, MCL’s scalability to cluster large datasets still remains a bottleneck due to high running times andmore » memory demands. In this paper, we present High-performance MCL (HipMCL), a parallel implementation of the original MCL algorithm that can run on distributed-memory computers. We show that HipMCL can efficiently utilize 2000 compute nodes and cluster a network of ~70 million nodes with ~68 billion edges in ~2.4 h. By exploiting distributed-memory environments, HipMCL clusters large-scale networks several orders of magnitude faster than MCL and enables clustering of even bigger networks. Finally, HipMCL is based on MPI and OpenMP and is freely available under a modified BSD license.« less

Elastic Cloud Computing Architecture and System for Heterogeneous Spatiotemporal Computing

NASA Astrophysics Data System (ADS)

Shi, X.

2017-10-01

Spatiotemporal computation implements a variety of different algorithms. When big data are involved, desktop computer or standalone application may not be able to complete the computation task due to limited memory and computing power. Now that a variety of hardware accelerators and computing platforms are available to improve the performance of geocomputation, different algorithms may have different behavior on different computing infrastructure and platforms. Some are perfect for implementation on a cluster of graphics processing units (GPUs), while GPUs may not be useful on certain kind of spatiotemporal computation. This is the same situation in utilizing a cluster of Intel's many-integrated-core (MIC) or Xeon Phi, as well as Hadoop or Spark platforms, to handle big spatiotemporal data. Furthermore, considering the energy efficiency requirement in general computation, Field Programmable Gate Array (FPGA) may be a better solution for better energy efficiency when the performance of computation could be similar or better than GPUs and MICs. It is expected that an elastic cloud computing architecture and system that integrates all of GPUs, MICs, and FPGAs could be developed and deployed to support spatiotemporal computing over heterogeneous data types and computational problems.

Scalable cloud without dedicated storage

NASA Astrophysics Data System (ADS)

Batkovich, D. V.; Kompaniets, M. V.; Zarochentsev, A. K.

2015-05-01

We present a prototype of a scalable computing cloud. It is intended to be deployed on the basis of a cluster without the separate dedicated storage. The dedicated storage is replaced by the distributed software storage. In addition, all cluster nodes are used both as computing nodes and as storage nodes. This solution increases utilization of the cluster resources as well as improves fault tolerance and performance of the distributed storage. Another advantage of this solution is high scalability with a relatively low initial and maintenance cost. The solution is built on the basis of the open source components like OpenStack, CEPH, etc.

OCCAM: a flexible, multi-purpose and extendable HPC cluster

NASA Astrophysics Data System (ADS)

Aldinucci, M.; Bagnasco, S.; Lusso, S.; Pasteris, P.; Rabellino, S.; Vallero, S.

2017-10-01

The Open Computing Cluster for Advanced data Manipulation (OCCAM) is a multipurpose flexible HPC cluster designed and operated by a collaboration between the University of Torino and the Sezione di Torino of the Istituto Nazionale di Fisica Nucleare. It is aimed at providing a flexible, reconfigurable and extendable infrastructure to cater to a wide range of different scientific computing use cases, including ones from solid-state chemistry, high-energy physics, computer science, big data analytics, computational biology, genomics and many others. Furthermore, it will serve as a platform for R&D activities on computational technologies themselves, with topics ranging from GPU acceleration to Cloud Computing technologies. A heterogeneous and reconfigurable system like this poses a number of challenges related to the frequency at which heterogeneous hardware resources might change their availability and shareability status, which in turn affect methods and means to allocate, manage, optimize, bill, monitor VMs, containers, virtual farms, jobs, interactive bare-metal sessions, etc. This work describes some of the use cases that prompted the design and construction of the HPC cluster, its architecture and resource provisioning model, along with a first characterization of its performance by some synthetic benchmark tools and a few realistic use-case tests.

Balancing computation and communication power in power constrained clusters

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

Piga, Leonardo; Paul, Indrani; Huang, Wei

Systems, apparatuses, and methods for balancing computation and communication power in power constrained environments. A data processing cluster with a plurality of compute nodes may perform parallel processing of a workload in a power constrained environment. Nodes that finish tasks early may be power-gated based on one or more conditions. In some scenarios, a node may predict a wait duration and go into a reduced power consumption state if the wait duration is predicted to be greater than a threshold. The power saved by power-gating one or more nodes may be reassigned for use by other nodes. A cluster agentmore » may be configured to reassign the unused power to the active nodes to expedite workload processing.« less

A uniform approach for programming distributed heterogeneous computing systems

PubMed Central

Grasso, Ivan; Pellegrini, Simone; Cosenza, Biagio; Fahringer, Thomas

2014-01-01

Large-scale compute clusters of heterogeneous nodes equipped with multi-core CPUs and GPUs are getting increasingly popular in the scientific community. However, such systems require a combination of different programming paradigms making application development very challenging. In this article we introduce libWater, a library-based extension of the OpenCL programming model that simplifies the development of heterogeneous distributed applications. libWater consists of a simple interface, which is a transparent abstraction of the underlying distributed architecture, offering advanced features such as inter-context and inter-node device synchronization. It provides a runtime system which tracks dependency information enforced by event synchronization to dynamically build a DAG of commands, on which we automatically apply two optimizations: collective communication pattern detection and device-host-device copy removal. We assess libWater’s performance in three compute clusters available from the Vienna Scientific Cluster, the Barcelona Supercomputing Center and the University of Innsbruck, demonstrating improved performance and scaling with different test applications and configurations. PMID:25844015

A uniform approach for programming distributed heterogeneous computing systems.

PubMed

Grasso, Ivan; Pellegrini, Simone; Cosenza, Biagio; Fahringer, Thomas

2014-12-01

Large-scale compute clusters of heterogeneous nodes equipped with multi-core CPUs and GPUs are getting increasingly popular in the scientific community. However, such systems require a combination of different programming paradigms making application development very challenging. In this article we introduce libWater, a library-based extension of the OpenCL programming model that simplifies the development of heterogeneous distributed applications. libWater consists of a simple interface, which is a transparent abstraction of the underlying distributed architecture, offering advanced features such as inter-context and inter-node device synchronization. It provides a runtime system which tracks dependency information enforced by event synchronization to dynamically build a DAG of commands, on which we automatically apply two optimizations: collective communication pattern detection and device-host-device copy removal. We assess libWater's performance in three compute clusters available from the Vienna Scientific Cluster, the Barcelona Supercomputing Center and the University of Innsbruck, demonstrating improved performance and scaling with different test applications and configurations.

Workload Characterization of a Leadership Class Storage Cluster

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

Kim, Youngjae; Gunasekaran, Raghul; Shipman, Galen M

2010-01-01

Understanding workload characteristics is critical for optimizing and improving the performance of current systems and software, and architecting new storage systems based on observed workload patterns. In this paper, we characterize the scientific workloads of the world s fastest HPC (High Performance Computing) storage cluster, Spider, at the Oak Ridge Leadership Computing Facility (OLCF). Spider provides an aggregate bandwidth of over 240 GB/s with over 10 petabytes of RAID 6 formatted capacity. OLCFs flagship petascale simulation platform, Jaguar, and other large HPC clusters, in total over 250 thousands compute cores, depend on Spider for their I/O needs. We characterize themore » system utilization, the demands of reads and writes, idle time, and the distribution of read requests to write requests for the storage system observed over a period of 6 months. From this study we develop synthesized workloads and we show that the read and write I/O bandwidth usage as well as the inter-arrival time of requests can be modeled as a Pareto distribution.« less

Spiking neural networks on high performance computer clusters

NASA Astrophysics Data System (ADS)

Chen, Chong; Taha, Tarek M.

2011-09-01

In this paper we examine the acceleration of two spiking neural network models on three clusters of multicore processors representing three categories of processors: x86, STI Cell, and NVIDIA GPGPUs. The x86 cluster utilized consists of 352 dualcore AMD Opterons, the Cell cluster consists of 320 Sony Playstation 3s, while the GPGPU cluster contains 32 NVIDIA Tesla S1070 systems. The results indicate that the GPGPU platform can dominate in performance compared to the Cell and x86 platforms examined. From a cost perspective, the GPGPU is more expensive in terms of neuron/s throughput. If the cost of GPGPUs go down in the future, this platform will become very cost effective for these models.

Adaptive density trajectory cluster based on time and space distance

NASA Astrophysics Data System (ADS)

Liu, Fagui; Zhang, Zhijie

2017-10-01

There are some hotspot problems remaining in trajectory cluster for discovering mobile behavior regularity, such as the computation of distance between sub trajectories, the setting of parameter values in cluster algorithm and the uncertainty/boundary problem of data set. As a result, based on the time and space, this paper tries to define the calculation method of distance between sub trajectories. The significance of distance calculation for sub trajectories is to clearly reveal the differences in moving trajectories and to promote the accuracy of cluster algorithm. Besides, a novel adaptive density trajectory cluster algorithm is proposed, in which cluster radius is computed through using the density of data distribution. In addition, cluster centers and number are selected by a certain strategy automatically, and uncertainty/boundary problem of data set is solved by designed weighted rough c-means. Experimental results demonstrate that the proposed algorithm can perform the fuzzy trajectory cluster effectively on the basis of the time and space distance, and obtain the optimal cluster centers and rich cluster results information adaptably for excavating the features of mobile behavior in mobile and sociology network.

Performance of fusion algorithms for computer-aided detection and classification of mines in very shallow water obtained from testing in navy Fleet Battle Exercise-Hotel 2000

NASA Astrophysics Data System (ADS)

Ciany, Charles M.; Zurawski, William; Kerfoot, Ian

2001-10-01

The performance of Computer Aided Detection/Computer Aided Classification (CAD/CAC) Fusion algorithms on side-scan sonar images was evaluated using data taken at the Navy's's Fleet Battle Exercise-Hotel held in Panama City, Florida, in August 2000. A 2-of-3 binary fusion algorithm is shown to provide robust performance. The algorithm accepts the classification decisions and associated contact locations form three different CAD/CAC algorithms, clusters the contacts based on Euclidian distance, and then declares a valid target when a clustered contact is declared by at least 2 of the 3 individual algorithms. This simple binary fusion provided a 96 percent probability of correct classification at a false alarm rate of 0.14 false alarms per image per side. The performance represented a 3.8:1 reduction in false alarms over the best performing single CAD/CAC algorithm, with no loss in probability of correct classification.

Grid Computing Application for Brain Magnetic Resonance Image Processing

NASA Astrophysics Data System (ADS)

Valdivia, F.; Crépeault, B.; Duchesne, S.

2012-02-01

This work emphasizes the use of grid computing and web technology for automatic post-processing of brain magnetic resonance images (MRI) in the context of neuropsychiatric (Alzheimer's disease) research. Post-acquisition image processing is achieved through the interconnection of several individual processes into pipelines. Each process has input and output data ports, options and execution parameters, and performs single tasks such as: a) extracting individual image attributes (e.g. dimensions, orientation, center of mass), b) performing image transformations (e.g. scaling, rotation, skewing, intensity standardization, linear and non-linear registration), c) performing image statistical analyses, and d) producing the necessary quality control images and/or files for user review. The pipelines are built to perform specific sequences of tasks on the alphanumeric data and MRIs contained in our database. The web application is coded in PHP and allows the creation of scripts to create, store and execute pipelines and their instances either on our local cluster or on high-performance computing platforms. To run an instance on an external cluster, the web application opens a communication tunnel through which it copies the necessary files, submits the execution commands and collects the results. We present result on system tests for the processing of a set of 821 brain MRIs from the Alzheimer's Disease Neuroimaging Initiative study via a nonlinear registration pipeline composed of 10 processes. Our results show successful execution on both local and external clusters, and a 4-fold increase in performance if using the external cluster. However, the latter's performance does not scale linearly as queue waiting times and execution overhead increase with the number of tasks to be executed.

Reweighted mass center based object-oriented sparse subspace clustering for hyperspectral images

NASA Astrophysics Data System (ADS)

Zhai, Han; Zhang, Hongyan; Zhang, Liangpei; Li, Pingxiang

2016-10-01

Considering the inevitable obstacles faced by the pixel-based clustering methods, such as salt-and-pepper noise, high computational complexity, and the lack of spatial information, a reweighted mass center based object-oriented sparse subspace clustering (RMC-OOSSC) algorithm for hyperspectral images (HSIs) is proposed. First, the mean-shift segmentation method is utilized to oversegment the HSI to obtain meaningful objects. Second, a distance reweighted mass center learning model is presented to extract the representative and discriminative features for each object. Third, assuming that all the objects are sampled from a union of subspaces, it is natural to apply the SSC algorithm to the HSI. Faced with the high correlation among the hyperspectral objects, a weighting scheme is adopted to ensure that the highly correlated objects are preferred in the procedure of sparse representation, to reduce the representation errors. Two widely used hyperspectral datasets were utilized to test the performance of the proposed RMC-OOSSC algorithm, obtaining high clustering accuracies (overall accuracy) of 71.98% and 89.57%, respectively. The experimental results show that the proposed method clearly improves the clustering performance with respect to the other state-of-the-art clustering methods, and it significantly reduces the computational time.

Integration of High-Performance Computing into Cloud Computing Services

NASA Astrophysics Data System (ADS)

Vouk, Mladen A.; Sills, Eric; Dreher, Patrick

High-Performance Computing (HPC) projects span a spectrum of computer hardware implementations ranging from peta-flop supercomputers, high-end tera-flop facilities running a variety of operating systems and applications, to mid-range and smaller computational clusters used for HPC application development, pilot runs and prototype staging clusters. What they all have in common is that they operate as a stand-alone system rather than a scalable and shared user re-configurable resource. The advent of cloud computing has changed the traditional HPC implementation. In this article, we will discuss a very successful production-level architecture and policy framework for supporting HPC services within a more general cloud computing infrastructure. This integrated environment, called Virtual Computing Lab (VCL), has been operating at NC State since fall 2004. Nearly 8,500,000 HPC CPU-Hrs were delivered by this environment to NC State faculty and students during 2009. In addition, we present and discuss operational data that show that integration of HPC and non-HPC (or general VCL) services in a cloud can substantially reduce the cost of delivering cloud services (down to cents per CPU hour).

Remote control system for high-perfomance computer simulation of crystal growth by the PFC method

NASA Astrophysics Data System (ADS)

Pavlyuk, Evgeny; Starodumov, Ilya; Osipov, Sergei

2017-04-01

Modeling of crystallization process by the phase field crystal method (PFC) - one of the important directions of modern computational materials science. In this paper, the practical side of the computer simulation of the crystallization process by the PFC method is investigated. To solve problems using this method, it is necessary to use high-performance computing clusters, data storage systems and other often expensive complex computer systems. Access to such resources is often limited, unstable and accompanied by various administrative problems. In addition, the variety of software and settings of different computing clusters sometimes does not allow researchers to use unified program code. There is a need to adapt the program code for each configuration of the computer complex. The practical experience of the authors has shown that the creation of a special control system for computing with the possibility of remote use can greatly simplify the implementation of simulations and increase the performance of scientific research. In current paper we show the principal idea of such a system and justify its efficiency.

Reexamine structures and relative stability of medium-sized silicon clusters: Low-lying endohedral fullerene-like clusters Si 30-Si 38

NASA Astrophysics Data System (ADS)

Yoo, Soohaeng; Shao, Nan; Zeng, X. C.

2009-10-01

We report improved results of lowest-lying silicon clusters Si 30-Si 38. A large population of low-energy clusters are collected from previous searches by several research groups and the binding energies of these clusters are computed using density-functional theory (DFT) methods. Best candidates (isomers with high binding energies) are identified from the screening calculations. Additional constrained search is then performed for the best candidates using the basin-hopping method combined with DFT geometry optimization. The obtained low-lying clusters are classified according to binding energies computed using either the Perdew-Burke-Ernzerhof (PBE) functional or the Becke exchange and Lee-Yang-Parr correlation (BLYP) functional. We propose to rank low-lying clusters according to the mean PBE/BLYP binding energies in view that the PBE functional tends to give greater binding energies for more compact clusters whereas the BLYP functional tends to give greater binding energies for less compact clusters or clusters composed of small-sized magic-number clusters. Except for Si 30, the new search confirms again that medium-size silicon clusters Si 31-Si 38 constructed with proper fullerene cage motifs are most promising to be the lowest-energy structures.

Characteristics of airflow and particle deposition in COPD current smokers

NASA Astrophysics Data System (ADS)

Zou, Chunrui; Choi, Jiwoong; Haghighi, Babak; Choi, Sanghun; Hoffman, Eric A.; Lin, Ching-Long

2017-11-01

A recent imaging-based cluster analysis of computed tomography (CT) lung images in a chronic obstructive pulmonary disease (COPD) cohort identified four clusters, viz. disease sub-populations. Cluster 1 had relatively normal airway structures; Cluster 2 had wall thickening; Cluster 3 exhibited decreased wall thickness and luminal narrowing; Cluster 4 had a significant decrease of luminal diameter and a significant reduction of lung deformation, thus having relatively low pulmonary functions. To better understand the characteristics of airflow and particle deposition in these clusters, we performed computational fluid and particle dynamics analyses on representative cluster patients and healthy controls using CT-based airway models and subject-specific 3D-1D coupled boundary conditions. The results show that particle deposition in central airways of cluster 4 patients was noticeably increased especially with increasing particle size despite reduced vital capacity as compared to other clusters and healthy controls. This may be attributable in part to significant airway constriction in cluster 4. This study demonstrates the potential application of cluster-guided CFD analysis in disease populations. NIH Grants U01HL114494 and S10-RR022421, and FDA Grant U01FD005837.

Computational cluster validation for microarray data analysis: experimental assessment of Clest, Consensus Clustering, Figure of Merit, Gap Statistics and Model Explorer.

PubMed

Giancarlo, Raffaele; Scaturro, Davide; Utro, Filippo

2008-10-29

Inferring cluster structure in microarray datasets is a fundamental task for the so-called -omic sciences. It is also a fundamental question in Statistics, Data Analysis and Classification, in particular with regard to the prediction of the number of clusters in a dataset, usually established via internal validation measures. Despite the wealth of internal measures available in the literature, new ones have been recently proposed, some of them specifically for microarray data. We consider five such measures: Clest, Consensus (Consensus Clustering), FOM (Figure of Merit), Gap (Gap Statistics) and ME (Model Explorer), in addition to the classic WCSS (Within Cluster Sum-of-Squares) and KL (Krzanowski and Lai index). We perform extensive experiments on six benchmark microarray datasets, using both Hierarchical and K-means clustering algorithms, and we provide an analysis assessing both the intrinsic ability of a measure to predict the correct number of clusters in a dataset and its merit relative to the other measures. We pay particular attention both to precision and speed. Moreover, we also provide various fast approximation algorithms for the computation of Gap, FOM and WCSS. The main result is a hierarchy of those measures in terms of precision and speed, highlighting some of their merits and limitations not reported before in the literature. Based on our analysis, we draw several conclusions for the use of those internal measures on microarray data. We report the main ones. Consensus is by far the best performer in terms of predictive power and remarkably algorithm-independent. Unfortunately, on large datasets, it may be of no use because of its non-trivial computer time demand (weeks on a state of the art PC). FOM is the second best performer although, quite surprisingly, it may not be competitive in this scenario: it has essentially the same predictive power of WCSS but it is from 6 to 100 times slower in time, depending on the dataset. The approximation algorithms for the computation of FOM, Gap and WCSS perform very well, i.e., they are faster while still granting a very close approximation of FOM and WCSS. The approximation algorithm for the computation of Gap deserves to be singled-out since it has a predictive power far better than Gap, it is competitive with the other measures, but it is at least two order of magnitude faster in time with respect to Gap. Another important novel conclusion that can be drawn from our analysis is that all the measures we have considered show severe limitations on large datasets, either due to computational demand (Consensus, as already mentioned, Clest and Gap) or to lack of precision (all of the other measures, including their approximations). The software and datasets are available under the GNU GPL on the supplementary material web page.

Hybrid cloud and cluster computing paradigms for life science applications

PubMed Central

2010-01-01

Background Clouds and MapReduce have shown themselves to be a broadly useful approach to scientific computing especially for parallel data intensive applications. However they have limited applicability to some areas such as data mining because MapReduce has poor performance on problems with an iterative structure present in the linear algebra that underlies much data analysis. Such problems can be run efficiently on clusters using MPI leading to a hybrid cloud and cluster environment. This motivates the design and implementation of an open source Iterative MapReduce system Twister. Results Comparisons of Amazon, Azure, and traditional Linux and Windows environments on common applications have shown encouraging performance and usability comparisons in several important non iterative cases. These are linked to MPI applications for final stages of the data analysis. Further we have released the open source Twister Iterative MapReduce and benchmarked it against basic MapReduce (Hadoop) and MPI in information retrieval and life sciences applications. Conclusions The hybrid cloud (MapReduce) and cluster (MPI) approach offers an attractive production environment while Twister promises a uniform programming environment for many Life Sciences applications. Methods We used commercial clouds Amazon and Azure and the NSF resource FutureGrid to perform detailed comparisons and evaluations of different approaches to data intensive computing. Several applications were developed in MPI, MapReduce and Twister in these different environments. PMID:21210982

Hybrid cloud and cluster computing paradigms for life science applications.

PubMed

Qiu, Judy; Ekanayake, Jaliya; Gunarathne, Thilina; Choi, Jong Youl; Bae, Seung-Hee; Li, Hui; Zhang, Bingjing; Wu, Tak-Lon; Ruan, Yang; Ekanayake, Saliya; Hughes, Adam; Fox, Geoffrey

2010-12-21

Clouds and MapReduce have shown themselves to be a broadly useful approach to scientific computing especially for parallel data intensive applications. However they have limited applicability to some areas such as data mining because MapReduce has poor performance on problems with an iterative structure present in the linear algebra that underlies much data analysis. Such problems can be run efficiently on clusters using MPI leading to a hybrid cloud and cluster environment. This motivates the design and implementation of an open source Iterative MapReduce system Twister. Comparisons of Amazon, Azure, and traditional Linux and Windows environments on common applications have shown encouraging performance and usability comparisons in several important non iterative cases. These are linked to MPI applications for final stages of the data analysis. Further we have released the open source Twister Iterative MapReduce and benchmarked it against basic MapReduce (Hadoop) and MPI in information retrieval and life sciences applications. The hybrid cloud (MapReduce) and cluster (MPI) approach offers an attractive production environment while Twister promises a uniform programming environment for many Life Sciences applications. We used commercial clouds Amazon and Azure and the NSF resource FutureGrid to perform detailed comparisons and evaluations of different approaches to data intensive computing. Several applications were developed in MPI, MapReduce and Twister in these different environments.

GATE Monte Carlo simulation in a cloud computing environment

NASA Astrophysics Data System (ADS)

Rowedder, Blake Austin

The GEANT4-based GATE is a unique and powerful Monte Carlo (MC) platform, which provides a single code library allowing the simulation of specific medical physics applications, e.g. PET, SPECT, CT, radiotherapy, and hadron therapy. However, this rigorous yet flexible platform is used only sparingly in the clinic due to its lengthy calculation time. By accessing the powerful computational resources of a cloud computing environment, GATE's runtime can be significantly reduced to clinically feasible levels without the sizable investment of a local high performance cluster. This study investigated a reliable and efficient execution of GATE MC simulations using a commercial cloud computing services. Amazon's Elastic Compute Cloud was used to launch several nodes equipped with GATE. Job data was initially broken up on the local computer, then uploaded to the worker nodes on the cloud. The results were automatically downloaded and aggregated on the local computer for display and analysis. Five simulations were repeated for every cluster size between 1 and 20 nodes. Ultimately, increasing cluster size resulted in a decrease in calculation time that could be expressed with an inverse power model. Comparing the benchmark results to the published values and error margins indicated that the simulation results were not affected by the cluster size and thus that integrity of a calculation is preserved in a cloud computing environment. The runtime of a 53 minute long simulation was decreased to 3.11 minutes when run on a 20-node cluster. The ability to improve the speed of simulation suggests that fast MC simulations are viable for imaging and radiotherapy applications. With high power computing continuing to lower in price and accessibility, implementing Monte Carlo techniques with cloud computing for clinical applications will continue to become more attractive.

Assessment of gene order computing methods for Alzheimer's disease

PubMed Central

2013-01-01

Background Computational genomics of Alzheimer disease (AD), the most common form of senile dementia, is a nascent field in AD research. The field includes AD gene clustering by computing gene order which generates higher quality gene clustering patterns than most other clustering methods. However, there are few available gene order computing methods such as Genetic Algorithm (GA) and Ant Colony Optimization (ACO). Further, their performance in gene order computation using AD microarray data is not known. We thus set forth to evaluate the performances of current gene order computing methods with different distance formulas, and to identify additional features associated with gene order computation. Methods Using different distance formulas- Pearson distance and Euclidean distance, the squared Euclidean distance, and other conditions, gene orders were calculated by ACO and GA (including standard GA and improved GA) methods, respectively. The qualities of the gene orders were compared, and new features from the calculated gene orders were identified. Results Compared to the GA methods tested in this study, ACO fits the AD microarray data the best when calculating gene order. In addition, the following features were revealed: different distance formulas generated a different quality of gene order, and the commonly used Pearson distance was not the best distance formula when used with both GA and ACO methods for AD microarray data. Conclusion Compared with Pearson distance and Euclidean distance, the squared Euclidean distance generated the best quality gene order computed by GA and ACO methods. PMID:23369541

WinHPC System | High-Performance Computing | NREL

Science.gov Websites

System WinHPC System NREL's WinHPC system is a computing cluster running the Microsoft Windows operating system. It allows users to run jobs requiring a Windows environment such as ANSYS and MATLAB

Construction of the energy matrix for complex atoms. Part VIII: Hyperfine structure HPC calculations for terbium atom

NASA Astrophysics Data System (ADS)

Elantkowska, Magdalena; Ruczkowski, Jarosław; Sikorski, Andrzej; Dembczyński, Jerzy

2017-11-01

A parametric analysis of the hyperfine structure (hfs) for the even parity configurations of atomic terbium (Tb I) is presented in this work. We introduce the complete set of 4fN-core states in our high-performance computing (HPC) calculations. For calculations of the huge hyperfine structure matrix, requiring approximately 5000 hours when run on a single CPU, we propose the methods utilizing a personal computer cluster or, alternatively a cluster of Microsoft Azure virtual machines (VM). These methods give a factor 12 performance boost, enabling the calculations to complete in an acceptable time.

Surprising performance for vibrational frequencies of the distinguishable clusters with singles and doubles (DCSD) and MP2.5 approximations

NASA Astrophysics Data System (ADS)

Kesharwani, Manoj K.; Sylvetsky, Nitai; Martin, Jan M. L.

2017-11-01

We show that the DCSD (distinguishable clusters with all singles and doubles) correlation method permits the calculation of vibrational spectra at near-CCSD(T) quality but at no more than CCSD cost, and with comparatively inexpensive analytical gradients. For systems dominated by a single reference configuration, even MP2.5 is a viable alternative, at MP3 cost. MP2.5 performance for vibrational frequencies is comparable to double hybrids such as DSD-PBEP86-D3BJ, but without resorting to empirical parameters. DCSD is also quite suitable for computing zero-point vibrational energies in computational thermochemistry.

Research on elastic resource management for multi-queue under cloud computing environment

NASA Astrophysics Data System (ADS)

CHENG, Zhenjing; LI, Haibo; HUANG, Qiulan; Cheng, Yaodong; CHEN, Gang

2017-10-01

As a new approach to manage computing resource, virtualization technology is more and more widely applied in the high-energy physics field. A virtual computing cluster based on Openstack was built at IHEP, using HTCondor as the job queue management system. In a traditional static cluster, a fixed number of virtual machines are pre-allocated to the job queue of different experiments. However this method cannot be well adapted to the volatility of computing resource requirements. To solve this problem, an elastic computing resource management system under cloud computing environment has been designed. This system performs unified management of virtual computing nodes on the basis of job queue in HTCondor based on dual resource thresholds as well as the quota service. A two-stage pool is designed to improve the efficiency of resource pool expansion. This paper will present several use cases of the elastic resource management system in IHEPCloud. The practical run shows virtual computing resource dynamically expanded or shrunk while computing requirements change. Additionally, the CPU utilization ratio of computing resource was significantly increased when compared with traditional resource management. The system also has good performance when there are multiple condor schedulers and multiple job queues.

Comparing the performance of biomedical clustering methods.

PubMed

Wiwie, Christian; Baumbach, Jan; Röttger, Richard

2015-11-01

Identifying groups of similar objects is a popular first step in biomedical data analysis, but it is error-prone and impossible to perform manually. Many computational methods have been developed to tackle this problem. Here we assessed 13 well-known methods using 24 data sets ranging from gene expression to protein domains. Performance was judged on the basis of 13 common cluster validity indices. We developed a clustering analysis platform, ClustEval (http://clusteval.mpi-inf.mpg.de), to promote streamlined evaluation, comparison and reproducibility of clustering results in the future. This allowed us to objectively evaluate the performance of all tools on all data sets with up to 1,000 different parameter sets each, resulting in a total of more than 4 million calculated cluster validity indices. We observed that there was no universal best performer, but on the basis of this wide-ranging comparison we were able to develop a short guideline for biomedical clustering tasks. ClustEval allows biomedical researchers to pick the appropriate tool for their data type and allows method developers to compare their tool to the state of the art.

Experiences using OpenMP based on Computer Directed Software DSM on a PC Cluster

NASA Technical Reports Server (NTRS)

Hess, Matthias; Jost, Gabriele; Mueller, Matthias; Ruehle, Roland

2003-01-01

In this work we report on our experiences running OpenMP programs on a commodity cluster of PCs running a software distributed shared memory (DSM) system. We describe our test environment and report on the performance of a subset of the NAS Parallel Benchmarks that have been automaticaly parallelized for OpenMP. We compare the performance of the OpenMP implementations with that of their message passing counterparts and discuss performance differences.

Implementing Simulation Design of Experiments and Remote Execution on a High Performance Computing Cluster

DTIC Science & Technology

2007-09-01

example, an application developed in Sun’s Netbeans [2007] integrated development environment (IDE) uses Swing class object for graphical user... Netbeans Version 5.5.1 [Computer Software]. Santa Clara, CA: Sun Microsystems. Process Modeler Version 7.0 [Computer Software]. Santa Clara, Ca

Continuous Variable Cluster State Generation over the Optical Spatial Mode Comb

DOE PAGES

Pooser, Raphael C.; Jing, Jietai

2014-10-20

One way quantum computing uses single qubit projective measurements performed on a cluster state (a highly entangled state of multiple qubits) in order to enact quantum gates. The model is promising due to its potential scalability; the cluster state may be produced at the beginning of the computation and operated on over time. Continuous variables (CV) offer another potential benefit in the form of deterministic entanglement generation. This determinism can lead to robust cluster states and scalable quantum computation. Recent demonstrations of CV cluster states have made great strides on the path to scalability utilizing either time or frequency multiplexingmore » in optical parametric oscillators (OPO) both above and below threshold. The techniques relied on a combination of entangling operators and beam splitter transformations. Here we show that an analogous transformation exists for amplifiers with Gaussian inputs states operating on multiple spatial modes. By judicious selection of local oscillators (LOs), the spatial mode distribution is analogous to the optical frequency comb consisting of axial modes in an OPO cavity. We outline an experimental system that generates cluster states across the spatial frequency comb which can also scale the amount of quantum noise reduction to potentially larger than in other systems.« less

Overlapping Community Detection based on Network Decomposition

NASA Astrophysics Data System (ADS)

Ding, Zhuanlian; Zhang, Xingyi; Sun, Dengdi; Luo, Bin

2016-04-01

Community detection in complex network has become a vital step to understand the structure and dynamics of networks in various fields. However, traditional node clustering and relatively new proposed link clustering methods have inherent drawbacks to discover overlapping communities. Node clustering is inadequate to capture the pervasive overlaps, while link clustering is often criticized due to the high computational cost and ambiguous definition of communities. So, overlapping community detection is still a formidable challenge. In this work, we propose a new overlapping community detection algorithm based on network decomposition, called NDOCD. Specifically, NDOCD iteratively splits the network by removing all links in derived link communities, which are identified by utilizing node clustering technique. The network decomposition contributes to reducing the computation time and noise link elimination conduces to improving the quality of obtained communities. Besides, we employ node clustering technique rather than link similarity measure to discover link communities, thus NDOCD avoids an ambiguous definition of community and becomes less time-consuming. We test our approach on both synthetic and real-world networks. Results demonstrate the superior performance of our approach both in computation time and accuracy compared to state-of-the-art algorithms.

Combining self-organizing mapping and supervised affinity propagation clustering approach to investigate functional brain networks involved in motor imagery and execution with fMRI measurements.

PubMed

Zhang, Jiang; Liu, Qi; Chen, Huafu; Yuan, Zhen; Huang, Jin; Deng, Lihua; Lu, Fengmei; Zhang, Junpeng; Wang, Yuqing; Wang, Mingwen; Chen, Liangyin

2015-01-01

Clustering analysis methods have been widely applied to identifying the functional brain networks of a multitask paradigm. However, the previously used clustering analysis techniques are computationally expensive and thus impractical for clinical applications. In this study a novel method, called SOM-SAPC that combines self-organizing mapping (SOM) and supervised affinity propagation clustering (SAPC), is proposed and implemented to identify the motor execution (ME) and motor imagery (MI) networks. In SOM-SAPC, SOM was first performed to process fMRI data and SAPC is further utilized for clustering the patterns of functional networks. As a result, SOM-SAPC is able to significantly reduce the computational cost for brain network analysis. Simulation and clinical tests involving ME and MI were conducted based on SOM-SAPC, and the analysis results indicated that functional brain networks were clearly identified with different response patterns and reduced computational cost. In particular, three activation clusters were clearly revealed, which include parts of the visual, ME and MI functional networks. These findings validated that SOM-SAPC is an effective and robust method to analyze the fMRI data with multitasks.

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.

FPGA cluster for high-performance AO real-time control system

NASA Astrophysics Data System (ADS)

Geng, Deli; Goodsell, Stephen J.; Basden, Alastair G.; Dipper, Nigel A.; Myers, Richard M.; Saunter, Chris D.

2006-06-01

Whilst the high throughput and low latency requirements for the next generation AO real-time control systems have posed a significant challenge to von Neumann architecture processor systems, the Field Programmable Gate Array (FPGA) has emerged as a long term solution with high performance on throughput and excellent predictability on latency. Moreover, FPGA devices have highly capable programmable interfacing, which lead to more highly integrated system. Nevertheless, a single FPGA is still not enough: multiple FPGA devices need to be clustered to perform the required subaperture processing and the reconstruction computation. In an AO real-time control system, the memory bandwidth is often the bottleneck of the system, simply because a vast amount of supporting data, e.g. pixel calibration maps and the reconstruction matrix, need to be accessed within a short period. The cluster, as a general computing architecture, has excellent scalability in processing throughput, memory bandwidth, memory capacity, and communication bandwidth. Problems, such as task distribution, node communication, system verification, are discussed.

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.

An open source software for fast grid-based data-mining in spatial epidemiology (FGBASE).

PubMed

Baker, David M; Valleron, Alain-Jacques

2014-10-30

Examining whether disease cases are clustered in space is an important part of epidemiological research. Another important part of spatial epidemiology is testing whether patients suffering from a disease are more, or less, exposed to environmental factors of interest than adequately defined controls. Both approaches involve determining the number of cases and controls (or population at risk) in specific zones. For cluster searches, this often must be done for millions of different zones. Doing this by calculating distances can lead to very lengthy computations. In this work we discuss the computational advantages of geographical grid-based methods, and introduce an open source software (FGBASE) which we have created for this purpose. Geographical grids based on the Lambert Azimuthal Equal Area projection are well suited for spatial epidemiology because they preserve area: each cell of the grid has the same area. We describe how data is projected onto such a grid, as well as grid-based algorithms for spatial epidemiological data-mining. The software program (FGBASE), that we have developed, implements these grid-based methods. The grid based algorithms perform extremely fast. This is particularly the case for cluster searches. When applied to a cohort of French Type 1 Diabetes (T1D) patients, as an example, the grid based algorithms detected potential clusters in a few seconds on a modern laptop. This compares very favorably to an equivalent cluster search using distance calculations instead of a grid, which took over 4 hours on the same computer. In the case study we discovered 4 potential clusters of T1D cases near the cities of Le Havre, Dunkerque, Toulouse and Nantes. One example of environmental analysis with our software was to study whether a significant association could be found between distance to vineyards with heavy pesticide. None was found. In both examples, the software facilitates the rapid testing of hypotheses. Grid-based algorithms for mining spatial epidemiological data provide advantages in terms of computational complexity thus improving the speed of computations. We believe that these methods and this software tool (FGBASE) will lower the computational barriers to entry for those performing epidemiological research.

General purpose molecular dynamics simulations fully implemented on graphics processing units

NASA Astrophysics Data System (ADS)

Anderson, Joshua A.; Lorenz, Chris D.; Travesset, A.

2008-05-01

Graphics processing units (GPUs), originally developed for rendering real-time effects in computer games, now provide unprecedented computational power for scientific applications. In this paper, we develop a general purpose molecular dynamics code that runs entirely on a single GPU. It is shown that our GPU implementation provides a performance equivalent to that of fast 30 processor core distributed memory cluster. Our results show that GPUs already provide an inexpensive alternative to such clusters and discuss implications for the future.

Pilot-in-the-Loop CFD Method Development

DTIC Science & Technology

2014-06-16

CFD analysis. Coupled simulations will be run at PSU on the COCOA -4 cluster, a high performance computing cluster. The CRUNCH CFD software has...been installed on the COCOA -4 servers and initial software tests are being conducted. Initial efforts will use the Generic Frigate Shape SFS-2 to

Utilizing the Structure and Content Information for XML Document Clustering

NASA Astrophysics Data System (ADS)

Tran, Tien; Kutty, Sangeetha; Nayak, Richi

This paper reports on the experiments and results of a clustering approach used in the INEX 2008 document mining challenge. The clustering approach utilizes both the structure and content information of the Wikipedia XML document collection. A latent semantic kernel (LSK) is used to measure the semantic similarity between XML documents based on their content features. The construction of a latent semantic kernel involves the computing of singular vector decomposition (SVD). On a large feature space matrix, the computation of SVD is very expensive in terms of time and memory requirements. Thus in this clustering approach, the dimension of the document space of a term-document matrix is reduced before performing SVD. The document space reduction is based on the common structural information of the Wikipedia XML document collection. The proposed clustering approach has shown to be effective on the Wikipedia collection in the INEX 2008 document mining challenge.

Development of a Computing Cluster At the University of Richmond

NASA Astrophysics Data System (ADS)

Carbonneau, J.; Gilfoyle, G. P.; Bunn, E. F.

2010-11-01

The University of Richmond has developed a computing cluster to support the massive simulation and data analysis requirements for programs in intermediate-energy nuclear physics, and cosmology. It is a 20-node, 240-core system running Red Hat Enterprise Linux 5. We have built and installed the physics software packages (Geant4, gemc, MADmap...) and developed shell and Perl scripts for running those programs on the remote nodes. The system has a theoretical processing peak of about 2500 GFLOPS. Testing with the High Performance Linpack (HPL) benchmarking program (one of the standard benchmarks used by the TOP500 list of fastest supercomputers) resulted in speeds of over 900 GFLOPS. The difference between the maximum and measured speeds is due to limitations in the communication speed among the nodes; creating a bottleneck for large memory problems. As HPL sends data between nodes, the gigabit Ethernet connection cannot keep up with the processing power. We will show how both the theoretical and actual performance of the cluster compares with other current and past clusters, as well as the cost per GFLOP. We will also examine the scaling of the performance when distributed to increasing numbers of nodes.

Distributed multitasking ITS with PVM

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

Fan, W.C.; Halbleib, J.A. Sr.

1995-12-31

Advances in computer hardware and communication software have made it possible to perform parallel-processing computing on a collection of desktop workstations. For many applications, multitasking on a cluster of high-performance workstations has achieved performance comparable to or better than that on a traditional supercomputer. From the point of view of cost-effectiveness, it also allows users to exploit available but unused computational resources and thus achieve a higher performance-to-cost ratio. Monte Carlo calculations are inherently parallelizable because the individual particle trajectories can be generated independently with minimum need for interprocessor communication. Furthermore, the number of particle histories that can be generatedmore » in a given amount of wall-clock time is nearly proportional to the number of processors in the cluster. This is an important fact because the inherent statistical uncertainty in any Monte Carlo result decreases as the number of histories increases. For these reasons, researchers have expended considerable effort to take advantage of different parallel architectures for a variety of Monte Carlo radiation transport codes, often with excellent results. The initial interest in this work was sparked by the multitasking capability of the MCNP code on a cluster of workstations using the Parallel Virtual Machine (PVM) software. On a 16-machine IBM RS/6000 cluster, it has been demonstrated that MCNP runs ten times as fast as on a single-processor CRAY YMP. In this paper, we summarize the implementation of a similar multitasking capability for the coupled electronphoton transport code system, the Integrated TIGER Series (ITS), and the evaluation of two load-balancing schemes for homogeneous and heterogeneous networks.« less

Distributed multitasking ITS with PVM

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

Fan, W.C.; Halbleib, J.A. Sr.

1995-02-01

Advances of computer hardware and communication software have made it possible to perform parallel-processing computing on a collection of desktop workstations. For many applications, multitasking on a cluster of high-performance workstations has achieved performance comparable or better than that on a traditional supercomputer. From the point of view of cost-effectiveness, it also allows users to exploit available but unused computational resources, and thus achieve a higher performance-to-cost ratio. Monte Carlo calculations are inherently parallelizable because the individual particle trajectories can be generated independently with minimum need for interprocessor communication. Furthermore, the number of particle histories that can be generated inmore » a given amount of wall-clock time is nearly proportional to the number of processors in the cluster. This is an important fact because the inherent statistical uncertainty in any Monte Carlo result decreases as the number of histories increases. For these reasons, researchers have expended considerable effort to take advantage of different parallel architectures for a variety of Monte Carlo radiation transport codes, often with excellent results. The initial interest in this work was sparked by the multitasking capability of MCNP on a cluster of workstations using the Parallel Virtual Machine (PVM) software. On a 16-machine IBM RS/6000 cluster, it has been demonstrated that MCNP runs ten times as fast as on a single-processor CRAY YMP. In this paper, we summarize the implementation of a similar multitasking capability for the coupled electron/photon transport code system, the Integrated TIGER Series (ITS), and the evaluation of two load balancing schemes for homogeneous and heterogeneous networks.« less

Experiences Using OpenMP Based on Compiler Directed Software DSM on a PC Cluster

NASA Technical Reports Server (NTRS)

Hess, Matthias; Jost, Gabriele; Mueller, Matthias; Ruehle, Roland; Biegel, Bryan (Technical Monitor)

2002-01-01

In this work we report on our experiences running OpenMP (message passing) programs on a commodity cluster of PCs (personal computers) running a software distributed shared memory (DSM) system. We describe our test environment and report on the performance of a subset of the NAS (NASA Advanced Supercomputing) Parallel Benchmarks that have been automatically parallelized for OpenMP. We compare the performance of the OpenMP implementations with that of their message passing counterparts and discuss performance differences.

Star Clusters Simulations Using GRAPE-5

NASA Astrophysics Data System (ADS)

Fukushige, Toshiyuki

We discuss simulations of star cluster, such as globular cluster, galaxy, and galaxy cluster, using GRAPE(GRAvity PipE)-5. GRAPE-5 is a new version of special-purpose computer for many-body simulation, GRAPE. GRAPE-5 has eight custom pipeline LSI (G5 chip) per board, and its peak performance is 38.4 Gflops. GRAPE-5 is different from its predecessor, GRAPE-3, regarding four points: a) the calculation speed per chip is 8 time faster, b) the PCI bus is adapted as an interface between host computer and GRAPE-5, and, therefore, the communication speed is order of magnitude faster, c) in addition to the pure 1/r potential, GRAPE-5 can calculate force with arbitrary cutoff function so that it can be applied to the Ewald or P3M methods, and d) the pair wise force calculated on GRAPE-5 is about 10 times more accurate. Using the GRAPE-5 system with Barnes-Hut tree algorithm, we can complete force calculations for one timestep in 10(N/106) seconds. This speed enables us to perform a pre-collapse globular cluster simulation with real number of particles, and a galaxy simulation with more than 1 million particles, within several days. We also present some results of star cluster simulations using the GRAPE-5 system.

Medical Imaging Lesion Detection Based on Unified Gravitational Fuzzy Clustering

PubMed Central

Vianney Kinani, Jean Marie; Gallegos Funes, Francisco; Mújica Vargas, Dante; Ramos Díaz, Eduardo; Arellano, Alfonso

2017-01-01

We develop a swift, robust, and practical tool for detecting brain lesions with minimal user intervention to assist clinicians and researchers in the diagnosis process, radiosurgery planning, and assessment of the patient's response to the therapy. We propose a unified gravitational fuzzy clustering-based segmentation algorithm, which integrates the Newtonian concept of gravity into fuzzy clustering. We first perform fuzzy rule-based image enhancement on our database which is comprised of T1/T2 weighted magnetic resonance (MR) and fluid-attenuated inversion recovery (FLAIR) images to facilitate a smoother segmentation. The scalar output obtained is fed into a gravitational fuzzy clustering algorithm, which separates healthy structures from the unhealthy. Finally, the lesion contour is automatically outlined through the initialization-free level set evolution method. An advantage of this lesion detection algorithm is its precision and its simultaneous use of features computed from the intensity properties of the MR scan in a cascading pattern, which makes the computation fast, robust, and self-contained. Furthermore, we validate our algorithm with large-scale experiments using clinical and synthetic brain lesion datasets. As a result, an 84%–93% overlap performance is obtained, with an emphasis on robustness with respect to different and heterogeneous types of lesion and a swift computation time. PMID:29158887

Computational Fluid Dynamics–Discrete Element Method (CFD-DEM) Study of Mass-Transfer Mechanisms in Riser Flow

PubMed Central

2017-01-01

We report a computational fluid dynamics–discrete element method (CFD-DEM) simulation study on the interplay between mass transfer and a heterogeneous catalyzed chemical reaction in cocurrent gas-particle flows as encountered in risers. Slip velocity, axial gas dispersion, gas bypassing, and particle mixing phenomena have been evaluated under riser flow conditions to study the complex system behavior in detail. The most important factors are found to be directly related to particle cluster formation. Low air-to-solids flux ratios lead to more heterogeneous systems, where the cluster formation is more pronounced and mass transfer more influenced. Falling clusters can be partially circumvented by the gas phase, which therefore does not fully interact with the cluster particles, leading to poor gas–solid contact efficiencies. Cluster gas–solid contact efficiencies are quantified at several gas superficial velocities, reaction rates, and dilution factors in order to gain more insight regarding the influence of clustering phenomena on the performance of riser reactors. PMID:28553011

Computational Fluid Dynamics-Discrete Element Method (CFD-DEM) Study of Mass-Transfer Mechanisms in Riser Flow.

PubMed

Carlos Varas, Álvaro E; Peters, E A J F; Kuipers, J A M

2017-05-17

We report a computational fluid dynamics-discrete element method (CFD-DEM) simulation study on the interplay between mass transfer and a heterogeneous catalyzed chemical reaction in cocurrent gas-particle flows as encountered in risers. Slip velocity, axial gas dispersion, gas bypassing, and particle mixing phenomena have been evaluated under riser flow conditions to study the complex system behavior in detail. The most important factors are found to be directly related to particle cluster formation. Low air-to-solids flux ratios lead to more heterogeneous systems, where the cluster formation is more pronounced and mass transfer more influenced. Falling clusters can be partially circumvented by the gas phase, which therefore does not fully interact with the cluster particles, leading to poor gas-solid contact efficiencies. Cluster gas-solid contact efficiencies are quantified at several gas superficial velocities, reaction rates, and dilution factors in order to gain more insight regarding the influence of clustering phenomena on the performance of riser reactors.

Architectural Principles and Experimentation of Distributed High Performance Virtual Clusters

ERIC Educational Resources Information Center

Younge, Andrew J.

2016-01-01

With the advent of virtualization and Infrastructure-as-a-Service (IaaS), the broader scientific computing community is considering the use of clouds for their scientific computing needs. This is due to the relative scalability, ease of use, advanced user environment customization abilities, and the many novel computing paradigms available for…

GPUs: An Emerging Platform for General-Purpose Computation

DTIC Science & Technology

2007-08-01

programming; real-time cinematic quality graphics Peak stream (26) License required (limited time no- cost evaluation program) Commercially...folding.stanford.edu (accessed 30 March 2007). 2. Fan, Z.; Qiu, F.; Kaufman, A.; Yoakum-Stover, S. GPU Cluster for High Performance Computing. ACM/IEEE...accessed 30 March 2007). 8. Goodnight, N.; Wang, R.; Humphreys, G. Computation on Programmable Graphics Hardware. IEEE Computer Graphics and

Plasma Physics Calculations on a Parallel Macintosh Cluster

NASA Astrophysics Data System (ADS)

Decyk, Viktor; Dauger, Dean; Kokelaar, Pieter

2000-03-01

We have constructed a parallel cluster consisting of 16 Apple Macintosh G3 computers running the MacOS, and achieved very good performance on numerically intensive, parallel plasma particle-in-cell simulations. A subset of the MPI message-passing library was implemented in Fortran77 and C. This library enabled us to port code, without modification, from other parallel processors to the Macintosh cluster. For large problems where message packets are large and relatively few in number, performance of 50-150 MFlops/node is possible, depending on the problem. This is fast enough that 3D calculations can be routinely done. Unlike Unix-based clusters, no special expertise in operating systems is required to build and run the cluster. Full details are available on our web site: http://exodus.physics.ucla.edu/appleseed/.

Plasma Physics Calculations on a Parallel Macintosh Cluster

NASA Astrophysics Data System (ADS)

Decyk, Viktor K.; Dauger, Dean E.; Kokelaar, Pieter R.

We have constructed a parallel cluster consisting of 16 Apple Macintosh G3 computers running the MacOS, and achieved very good performance on numerically intensive, parallel plasma particle-in-cell simulations. A subset of the MPI message-passing library was implemented in Fortran77 and C. This library enabled us to port code, without modification, from other parallel processors to the Macintosh cluster. For large problems where message packets are large and relatively few in number, performance of 50-150 Mflops/node is possible, depending on the problem. This is fast enough that 3D calculations can be routinely done. Unlike Unix-based clusters, no special expertise in operating systems is required to build and run the cluster. Full details are available on our web site: http://exodus.physics.ucla.edu/appleseed/.

Certification of Completion of Level-2 Milestone 464: Complete Phase 1 Integration of Site-Wide Global Parallel File System (SWGPFS)

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

Heidelberg, S T; Fitzgerald, K J; Richmond, G H

2006-01-24

There has been substantial development of the Lustre parallel filesystem prior to the configuration described below for this milestone. The initial Lustre filesystems that were deployed were directly connected to the cluster interconnect, i.e. Quadrics Elan3. That is, the clients (OSSes) and Meta-data Servers (MDS) were all directly connected to the cluster's internal high speed interconnect. This configuration serves a single cluster very well, but does not provide sharing of the filesystem among clusters. LLNL funded the development of high-efficiency ''portals router'' code by CFS (the company that develops Lustre) to enable us to move the Lustre servers to amore » GigE-connected network configuration, thus making it possible to connect to the servers from several clusters. With portals routing available, here is what changes: (1) another storage-only cluster is deployed to front the Lustre storage devices (these become the Lustre OSSes and MDS), (2) this ''Lustre cluster'' is attached via GigE connections to a large GigE switch/router cloud, (3) a small number of compute-cluster nodes are designated as ''gateway'' or ''portal router'' nodes, and (4) the portals router nodes are GigE-connected to the switch/router cloud. The Lustre configuration is then changed to reflect the new network paths. A typical example of this is a compute cluster and a related visualization cluster: the compute cluster produces the data (writes it to the Lustre filesystem), and the visualization cluster consumes some of the data (reads it from the Lustre filesystem). This process can be expanded by aggregating several collections of Lustre backend storage resources into one or more ''centralized'' Lustre filesystems, and then arranging to have several ''client'' clusters mount these centralized filesystems. The ''client clusters'' can be any combination of compute, visualization, archiving, or other types of cluster. This milestone demonstrates the operation and performance of a scaled-down version of such a large, centralized, shared Lustre filesystem concept.« less

Spectral gene set enrichment (SGSE).

PubMed

Frost, H Robert; Li, Zhigang; Moore, Jason H

2015-03-03

Gene set testing is typically performed in a supervised context to quantify the association between groups of genes and a clinical phenotype. In many cases, however, a gene set-based interpretation of genomic data is desired in the absence of a phenotype variable. Although methods exist for unsupervised gene set testing, they predominantly compute enrichment relative to clusters of the genomic variables with performance strongly dependent on the clustering algorithm and number of clusters. We propose a novel method, spectral gene set enrichment (SGSE), for unsupervised competitive testing of the association between gene sets and empirical data sources. SGSE first computes the statistical association between gene sets and principal components (PCs) using our principal component gene set enrichment (PCGSE) method. The overall statistical association between each gene set and the spectral structure of the data is then computed by combining the PC-level p-values using the weighted Z-method with weights set to the PC variance scaled by Tracy-Widom test p-values. Using simulated data, we show that the SGSE algorithm can accurately recover spectral features from noisy data. To illustrate the utility of our method on real data, we demonstrate the superior performance of the SGSE method relative to standard cluster-based techniques for testing the association between MSigDB gene sets and the variance structure of microarray gene expression data. Unsupervised gene set testing can provide important information about the biological signal held in high-dimensional genomic data sets. Because it uses the association between gene sets and samples PCs to generate a measure of unsupervised enrichment, the SGSE method is independent of cluster or network creation algorithms and, most importantly, is able to utilize the statistical significance of PC eigenvalues to ignore elements of the data most likely to represent noise.

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

Hadgu, Teklu; Appel, Gordon John

Sandia National Laboratories (SNL) continued evaluation of total system performance assessment (TSPA) computing systems for the previously considered Yucca Mountain Project (YMP). This was done to maintain the operational readiness of the computing infrastructure (computer hardware and software) and knowledge capability for total system performance assessment (TSPA) type analysis, as directed by the National Nuclear Security Administration (NNSA), DOE 2010. This work is a continuation of the ongoing readiness evaluation reported in Lee and Hadgu (2014) and Hadgu et al. (2015). The TSPA computing hardware (CL2014) and storage system described in Hadgu et al. (2015) were used for the currentmore » analysis. One floating license of GoldSim with Versions 9.60.300, 10.5 and 11.1.6 was installed on the cluster head node, and its distributed processing capability was mapped on the cluster processors. Other supporting software were tested and installed to support the TSPA-type analysis on the server cluster. The current tasks included verification of the TSPA-LA uncertainty and sensitivity analyses, and preliminary upgrade of the TSPA-LA from Version 9.60.300 to the latest version 11.1. All the TSPA-LA uncertainty and sensitivity analyses modeling cases were successfully tested and verified for the model reproducibility on the upgraded 2014 server cluster (CL2014). The uncertainty and sensitivity analyses used TSPA-LA modeling cases output generated in FY15 based on GoldSim Version 9.60.300 documented in Hadgu et al. (2015). The model upgrade task successfully converted the Nominal Modeling case to GoldSim Version 11.1. Upgrade of the remaining of the modeling cases and distributed processing tasks will continue. The 2014 server cluster and supporting software systems are fully operational to support TSPA-LA type analysis.« less

Sulfur in Cometary Dust

NASA Technical Reports Server (NTRS)

Fomenkova, M. N.

1997-01-01

The computer-intensive project consisted of the analysis and synthesis of existing data on composition of comet Halley dust particles. The main objective was to obtain a complete inventory of sulfur containing compounds in the comet Halley dust by building upon the existing classification of organic and inorganic compounds and applying a variety of statistical techniques for cluster and cross-correlational analyses. A student hired for this project wrote and tested the software to perform cluster analysis. The following tasks were carried out: (1) selecting the data from existing database for the proposed project; (2) finding access to a standard library of statistical routines for cluster analysis; (3) reformatting the data as necessary for input into the library routines; (4) performing cluster analysis and constructing hierarchical cluster trees using three methods to define the proximity of clusters; (5) presenting the output results in different formats to facilitate the interpretation of the obtained cluster trees; (6) selecting groups of data points common for all three trees as stable clusters. We have also considered the chemistry of sulfur in inorganic compounds.

GRAPE-6A: A Single-Card GRAPE-6 for Parallel PC-GRAPE Cluster Systems

NASA Astrophysics Data System (ADS)

Fukushige, Toshiyuki; Makino, Junichiro; Kawai, Atsushi

2005-12-01

In this paper, we describe the design and performance of GRAPE-6A, a special-purpose computer for gravitational many-body simulations. It was designed to be used with a PC cluster, in which each node has one GRAPE-6A. Such a configuration is particularly cost-effective in running parallel tree algorithms. Though the use of parallel tree algorithms was possible with the original GRAPE-6 hardware, it was not very cost-effective since a single GRAPE-6 board was still too fast and too expensive. Therefore, we designed GRAPE-6A as a single PCI card to minimize the reproduction cost and to optimize the computing speed. The peak performance is 130 Gflops for one GRAPE-6A board and 3.1 Tflops for our 24 node cluster. We describe the implementation of the tree, TreePM and individual timestep algorithms on both a single GRAPE-6A system and GRAPE-6A cluster. Using the tree algorithm on our 16-node GRAPE-6A system, we can complete a collisionless simulation with 100 million particles (8000 steps) within 10 days.

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

Finding Semirigid Domains in Biomolecules by Clustering Pair-Distance Variations

PubMed Central

Schreiner, Wolfgang

2014-01-01

Dynamic variations in the distances between pairs of atoms are used for clustering subdomains of biomolecules. We draw on a well-known target function for clustering and first show mathematically that the assignment of atoms to clusters has to be crisp, not fuzzy, as hitherto assumed. This reduces the computational load of clustering drastically, and we demonstrate results for several biomolecules relevant in immunoinformatics. Results are evaluated regarding the number of clusters, cluster size, cluster stability, and the evolution of clusters over time. Crisp clustering lends itself as an efficient tool to locate semirigid domains in the simulation of biomolecules. Such domains seem crucial for an optimum performance of subsequent statistical analyses, aiming at detecting minute motional patterns related to antigen recognition and signal transduction. PMID:24959586

A stellar audit: the computation of encounter rates for 47 Tucanae and omega Centauri

NASA Astrophysics Data System (ADS)

Davies, Melvyn B.; Benz, Willy

1995-10-01

Using King-Mitchie models, we compute encounter rates between the various stellar species in the globular clusters omega Cen and 47 Tuc. We also compute event rates for encounters between single stars and a population of primordial binaries. Using these rates, and what we have learnt from hydrodynamical simulations of encounters performed earlier, we compute the production rates of objects such as low-mass X-ray binaries (LMXBs), smothered neutron stars and blue stragglers (massive main-sequence stars). If 10 per cent of the stars are contained in primordial binaries, the production rate of interesting objects from encounters involving these binaries is as large as that from encounters between single stars. For example, encounters involving binaries produce a significant number of blue stragglers in both globular cluster models. The number of smothered neutron stars may exceed the number of LMXBs by a factor of 5-20, which may help to explain why millisecond pulsars are observed to outnumber LMXBs in globular clusters.

Quantum Computational Universality of the 2D Cai-Miyake-D"ur-Briegel Quantum State

NASA Astrophysics Data System (ADS)

Wei, Tzu-Chieh; Raussendorf, Robert; Kwek, Leong Chuan

2012-02-01

Universal quantum computation can be achieved by simply performing single-qubit measurements on a highly entangled resource state, such as cluster states. Cai, Miyake, D"ur, and Briegel recently constructed a ground state of a two-dimensional quantum magnet by combining multiple Affleck-Kennedy-Lieb-Tasaki quasichains of mixed spin-3/2 and spin-1/2 entities and by mapping pairs of neighboring spin-1/2 particles to individual spin-3/2 particles [Phys. Rev. A 82, 052309 (2010)]. They showed that this state enables universal quantum computation by constructing single- and two-qubit universal gates. Here, we give an alternative understanding of how this state gives rise to universal measurement-based quantum computation: by local operations, each quasichain can be converted to a one-dimensional cluster state and entangling gates between two neighboring logical qubits can be implemented by single-spin measurements. Furthermore, a two-dimensional cluster state can be distilled from the Cai-Miyake-D"ur-Briegel state.

Quantum computational universality of the Cai-Miyake-Duer-Briegel two-dimensional quantum state from Affleck-Kennedy-Lieb-Tasaki quasichains

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

Wei, Tzu-Chieh; C. N. Yang Institute for Theoretical Physics, State University of New York at Stony Brook, Stony Brook, New York 11794-3840; Raussendorf, Robert

2011-10-15

Universal quantum computation can be achieved by simply performing single-qubit measurements on a highly entangled resource state, such as cluster states. Cai, Miyake, Duer, and Briegel recently constructed a ground state of a two-dimensional quantum magnet by combining multiple Affleck-Kennedy-Lieb-Tasaki quasichains of mixed spin-3/2 and spin-1/2 entities and by mapping pairs of neighboring spin-1/2 particles to individual spin-3/2 particles [Phys. Rev. A 82, 052309 (2010)]. They showed that this state enables universal quantum computation by single-spin measurements. Here, we give an alternative understanding of how this state gives rise to universal measurement-based quantum computation: by local operations, each quasichain canmore » be converted to a one-dimensional cluster state and entangling gates between two neighboring logical qubits can be implemented by single-spin measurements. We further argue that a two-dimensional cluster state can be distilled from the Cai-Miyake-Duer-Briegel state.« less

Semisupervised Clustering by Iterative Partition and Regression with Neuroscience Applications

PubMed Central

Qian, Guoqi; Wu, Yuehua; Ferrari, Davide; Qiao, Puxue; Hollande, Frédéric

2016-01-01

Regression clustering is a mixture of unsupervised and supervised statistical learning and data mining method which is found in a wide range of applications including artificial intelligence and neuroscience. It performs unsupervised learning when it clusters the data according to their respective unobserved regression hyperplanes. The method also performs supervised learning when it fits regression hyperplanes to the corresponding data clusters. Applying regression clustering in practice requires means of determining the underlying number of clusters in the data, finding the cluster label of each data point, and estimating the regression coefficients of the model. In this paper, we review the estimation and selection issues in regression clustering with regard to the least squares and robust statistical methods. We also provide a model selection based technique to determine the number of regression clusters underlying the data. We further develop a computing procedure for regression clustering estimation and selection. Finally, simulation studies are presented for assessing the procedure, together with analyzing a real data set on RGB cell marking in neuroscience to illustrate and interpret the method. PMID:27212939

Cloud CPFP: a shotgun proteomics data analysis pipeline using cloud and high performance computing.

PubMed

Trudgian, David C; Mirzaei, Hamid

2012-12-07

We have extended the functionality of the Central Proteomics Facilities Pipeline (CPFP) to allow use of remote cloud and high performance computing (HPC) resources for shotgun proteomics data processing. CPFP has been modified to include modular local and remote scheduling for data processing jobs. The pipeline can now be run on a single PC or server, a local cluster, a remote HPC cluster, and/or the Amazon Web Services (AWS) cloud. We provide public images that allow easy deployment of CPFP in its entirety in the AWS cloud. This significantly reduces the effort necessary to use the software, and allows proteomics laboratories to pay for compute time ad hoc, rather than obtaining and maintaining expensive local server clusters. Alternatively the Amazon cloud can be used to increase the throughput of a local installation of CPFP as necessary. We demonstrate that cloud CPFP allows users to process data at higher speed than local installations but with similar cost and lower staff requirements. In addition to the computational improvements, the web interface to CPFP is simplified, and other functionalities are enhanced. The software is under active development at two leading institutions and continues to be released under an open-source license at http://cpfp.sourceforge.net.

Time-efficient simulations of tight-binding electronic structures with Intel Xeon PhiTM many-core processors

NASA Astrophysics Data System (ADS)

Ryu, Hoon; Jeong, Yosang; Kang, Ji-Hoon; Cho, Kyu Nam

2016-12-01

Modelling of multi-million atomic semiconductor structures is important as it not only predicts properties of physically realizable novel materials, but can accelerate advanced device designs. This work elaborates a new Technology-Computer-Aided-Design (TCAD) tool for nanoelectronics modelling, which uses a sp3d5s∗ tight-binding approach to describe multi-million atomic structures, and simulate electronic structures with high performance computing (HPC), including atomic effects such as alloy and dopant disorders. Being named as Quantum simulation tool for Advanced Nanoscale Devices (Q-AND), the tool shows nice scalability on traditional multi-core HPC clusters implying the strong capability of large-scale electronic structure simulations, particularly with remarkable performance enhancement on latest clusters of Intel Xeon PhiTM coprocessors. A review of the recent modelling study conducted to understand an experimental work of highly phosphorus-doped silicon nanowires, is presented to demonstrate the utility of Q-AND. Having been developed via Intel Parallel Computing Center project, Q-AND will be open to public to establish a sound framework of nanoelectronics modelling with advanced HPC clusters of a many-core base. With details of the development methodology and exemplary study of dopant electronics, this work will present a practical guideline for TCAD development to researchers in the field of computational nanoelectronics.

Scalable cluster administration - Chiba City I approach and lessons learned.

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

Navarro, J. P.; Evard, R.; Nurmi, D.

2002-07-01

Systems administrators of large clusters often need to perform the same administrative activity hundreds or thousands of times. Often such activities are time-consuming, especially the tasks of installing and maintaining software. By combining network services such as DHCP, TFTP, FTP, HTTP, and NFS with remote hardware control, cluster administrators can automate all administrative tasks. Scalable cluster administration addresses the following challenge: What systems design techniques can cluster builders use to automate cluster administration on very large clusters? We describe the approach used in the Mathematics and Computer Science Division of Argonne National Laboratory on Chiba City I, a 314-node Linuxmore » cluster; and we analyze the scalability, flexibility, and reliability benefits and limitations from that approach.« less

Effect of Graphene with Nanopores on Metal Clusters

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

Zhou, Hu; Chen, Xianlang; Wang, Lei

Porous graphene, which is a novel type of defective graphene, shows excellent potential as a support material for metal clusters. In this work, the stability and electronic structures of metal clusters (Pd, Ir, Rh) supported on pristine graphene and graphene with different sizes of nanopore were investigated by first-principle density functional theory (DFT) calculations. Thereafter, CO adsorption and oxidation reaction on the Pd-graphene system were chosen to evaluate its catalytic performance. Graphene with nanopore can strongly stabilize the metal clusters and cause a substantial downshift of the d-band center of the metal clusters, thus decreasing CO adsorption. All binding energies,more » d-band centers, and adsorption energies show a linear change with the size of the nanopore: a bigger size of nanopore corresponds to a stronger metal clusters bond to the graphene, lower downshift of the d-band center, and weaker CO adsorption. By using a suitable size nanopore, supported Pd clusters on the graphene will have similar CO and O2 adsorption ability, thus leading to superior CO tolerance. The DFT calculated reaction energy barriers show that graphene with nanopore is a superior catalyst for CO oxidation reaction. These properties can play an important role in instructing graphene-supported metal catalyst preparation to prevent the diffusion or agglomeration of metal clusters and enhance catalytic performance. This work was supported by National Basic Research Program of China (973Program) (2013CB733501), the National Natural Science Foundation of China (NSFC-21176221, 21136001, 21101137, 21306169, and 91334013). D. Mei acknowledges the support from the US Department of Energy, Office of Science, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences & Biosciences. Pacific Northwest National Laboratory (PNNL) is a multiprogram national laboratory operated for DOE by Battelle. Computing time was granted by the grand challenge of computational catalysis of the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL) and by the National Energy Research Scientific Computing Center (NERSC).« less

High Performance Computing (HPC) Innovation Service Portal Pilots Cloud Computing (HPC-ISP Pilot Cloud Computing)

DTIC Science & Technology

2011-08-01

5 Figure 4 Architetural diagram of running Blender on Amazon EC2 through Nimbis...classification of streaming data. Example input images (top left). All digit prototypes (cluster centers) found, with size proportional to frequency (top...Figure 4 Architetural diagram of running Blender on Amazon EC2 through Nimbis 1 http

Micromagnetics on high-performance workstation and mobile computational platforms

NASA Astrophysics Data System (ADS)

Fu, S.; Chang, R.; Couture, S.; Menarini, M.; Escobar, M. A.; Kuteifan, M.; Lubarda, M.; Gabay, D.; Lomakin, V.

2015-05-01

The feasibility of using high-performance desktop and embedded mobile computational platforms is presented, including multi-core Intel central processing unit, Nvidia desktop graphics processing units, and Nvidia Jetson TK1 Platform. FastMag finite element method-based micromagnetic simulator is used as a testbed, showing high efficiency on all the platforms. Optimization aspects of improving the performance of the mobile systems are discussed. The high performance, low cost, low power consumption, and rapid performance increase of the embedded mobile systems make them a promising candidate for micromagnetic simulations. Such architectures can be used as standalone systems or can be built as low-power computing clusters.

Stable isomers and electronic, vibrational, and optical properties of WS2 nano-clusters: A first-principles study

NASA Astrophysics Data System (ADS)

Hafizi, Roohollah; Hashemifar, S. Javad; Alaei, Mojtaba; Jangrouei, MohammadReza; Akbarzadeh, Hadi

2016-12-01

In this paper, we employ an evolutionary algorithm along with the full-potential density functional theory (DFT) computations to perform a comprehensive search for the stable structures of stoichiometric (WS2)n nano-clusters (n = 1 - 9), within three different exchange-correlation functionals. Our results suggest that n = 5 and 8 are possible candidates for the low temperature magic sizes of WS2 nano-clusters while at temperatures above 500 Kelvin, n = 7 exhibits a comparable relative stability with n = 8. The electronic properties and energy gap of the lowest energy isomers were computed within several schemes, including semilocal Perdew-Burke-Ernzerhof and Becke-Lee-Yang-Parr functionals, hybrid B3LYP functional, many body based DFT+GW approach, ΔSCF method, and time dependent DFT calculations. Vibrational spectra of the lowest lying isomers, computed by the force constant method, are used to address IR spectra and thermal free energy of the clusters. Time dependent density functional calculation in a real time domain is applied to determine the full absorption spectra and optical gap of the lowest energy isomers of the WS2 nano-clusters.

Software for Brain Network Simulations: A Comparative Study

PubMed Central

Tikidji-Hamburyan, Ruben A.; Narayana, Vikram; Bozkus, Zeki; El-Ghazawi, Tarek A.

2017-01-01

Numerical simulations of brain networks are a critical part of our efforts in understanding brain functions under pathological and normal conditions. For several decades, the community has developed many software packages and simulators to accelerate research in computational neuroscience. In this article, we select the three most popular simulators, as determined by the number of models in the ModelDB database, such as NEURON, GENESIS, and BRIAN, and perform an independent evaluation of these simulators. In addition, we study NEST, one of the lead simulators of the Human Brain Project. First, we study them based on one of the most important characteristics, the range of supported models. Our investigation reveals that brain network simulators may be biased toward supporting a specific set of models. However, all simulators tend to expand the supported range of models by providing a universal environment for the computational study of individual neurons and brain networks. Next, our investigations on the characteristics of computational architecture and efficiency indicate that all simulators compile the most computationally intensive procedures into binary code, with the aim of maximizing their computational performance. However, not all simulators provide the simplest method for module development and/or guarantee efficient binary code. Third, a study of their amenability for high-performance computing reveals that NEST can almost transparently map an existing model on a cluster or multicore computer, while NEURON requires code modification if the model developed for a single computer has to be mapped on a computational cluster. Interestingly, parallelization is the weakest characteristic of BRIAN, which provides no support for cluster computations and limited support for multicore computers. Fourth, we identify the level of user support and frequency of usage for all simulators. Finally, we carry out an evaluation using two case studies: a large network with simplified neural and synaptic models and a small network with detailed models. These two case studies allow us to avoid any bias toward a particular software package. The results indicate that BRIAN provides the most concise language for both cases considered. Furthermore, as expected, NEST mostly favors large network models, while NEURON is better suited for detailed models. Overall, the case studies reinforce our general observation that simulators have a bias in the computational performance toward specific types of the brain network models. PMID:28775687

Distributed computing for membrane-based modeling of action potential propagation.

PubMed

Porras, D; Rogers, J M; Smith, W M; Pollard, A E

2000-08-01

Action potential propagation simulations with physiologic membrane currents and macroscopic tissue dimensions are computationally expensive. We, therefore, analyzed distributed computing schemes to reduce execution time in workstation clusters by parallelizing solutions with message passing. Four schemes were considered in two-dimensional monodomain simulations with the Beeler-Reuter membrane equations. Parallel speedups measured with each scheme were compared to theoretical speedups, recognizing the relationship between speedup and code portions that executed serially. A data decomposition scheme based on total ionic current provided the best performance. Analysis of communication latencies in that scheme led to a load-balancing algorithm in which measured speedups at 89 +/- 2% and 75 +/- 8% of theoretical speedups were achieved in homogeneous and heterogeneous clusters of workstations. Speedups in this scheme with the Luo-Rudy dynamic membrane equations exceeded 3.0 with eight distributed workstations. Cluster speedups were comparable to those measured during parallel execution on a shared memory machine.

FY17 Status Report on the Computing Systems for the Yucca Mountain Project TSPA-LA Models.

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

Appel, Gordon John; Hadgu, Teklu; Appel, Gordon John

Sandia National Laboratories (SNL) continued evaluation of total system performance assessment (TSPA) computing systems for the previously considered Yucca Mountain Project (YMP). This was done to maintain the operational readiness of the computing infrastructure (computer hardware and software) and knowledge capability for total system performance assessment (TSPA) type analysis, as directed by the National Nuclear Security Administration (NNSA), DOE 2010. This work is a continuation of the ongoing readiness evaluation reported in Lee and Hadgu (2014), Hadgu et al. (2015) and Hadgu and Appel (2016). The TSPA computing hardware (CL2014) and storage system described in Hadgu et al. (2015) weremore » used for the current analysis. One floating license of GoldSim with Versions 9.60.300, 10.5, 11.1 and 12.0 was installed on the cluster head node, and its distributed processing capability was mapped on the cluster processors. Other supporting software were tested and installed to support the TSPA- type analysis on the server cluster. The current tasks included preliminary upgrade of the TSPA-LA from Version 9.60.300 to the latest version 12.0 and address DLL-related issues observed in the FY16 work. The model upgrade task successfully converted the Nominal Modeling case to GoldSim Versions 11.1/12. Conversions of the rest of the TSPA models were also attempted but program and operational difficulties precluded this. Upgrade of the remaining of the modeling cases and distributed processing tasks is expected to continue. The 2014 server cluster and supporting software systems are fully operational to support TSPA-LA type analysis.« less

HPC enabled real-time remote processing of laparoscopic surgery

NASA Astrophysics Data System (ADS)

Ronaghi, Zahra; Sapra, Karan; Izard, Ryan; Duffy, Edward; Smith, Melissa C.; Wang, Kuang-Ching; Kwartowitz, David M.

2016-03-01

Laparoscopic surgery is a minimally invasive surgical technique. The benefit of small incisions has a disadvantage of limited visualization of subsurface tissues. Image-guided surgery (IGS) uses pre-operative and intra-operative images to map subsurface structures. One particular laparoscopic system is the daVinci-si robotic surgical system. The video streams generate approximately 360 megabytes of data per second. Real-time processing this large stream of data on a bedside PC, single or dual node setup, has become challenging and a high-performance computing (HPC) environment may not always be available at the point of care. To process this data on remote HPC clusters at the typical 30 frames per second rate, it is required that each 11.9 MB video frame be processed by a server and returned within 1/30th of a second. We have implement and compared performance of compression, segmentation and registration algorithms on Clemson's Palmetto supercomputer using dual NVIDIA K40 GPUs per node. Our computing framework will also enable reliability using replication of computation. We will securely transfer the files to remote HPC clusters utilizing an OpenFlow-based network service, Steroid OpenFlow Service (SOS) that can increase performance of large data transfers over long-distance and high bandwidth networks. As a result, utilizing high-speed OpenFlow- based network to access computing clusters with GPUs will improve surgical procedures by providing real-time medical image processing and laparoscopic data.

Benchmark CCSD(T) and DFT study of binding energies in Be7 - 12: in search of reliable DFT functional for beryllium clusters

NASA Astrophysics Data System (ADS)

Labanc, Daniel; Šulka, Martin; Pitoňák, Michal; Černušák, Ivan; Urban, Miroslav; Neogrády, Pavel

2018-05-01

We present a computational study of the stability of small homonuclear beryllium clusters Be7 - 12 in singlet electronic states. Our predictions are based on highly correlated CCSD(T) coupled cluster calculations. Basis set convergence towards the complete basis set limit as well as the role of the 1s core electron correlation are carefully examined. Our CCSD(T) data for binding energies of Be7 - 12 clusters serve as a benchmark for performance assessment of several density functional theory (DFT) methods frequently used in beryllium cluster chemistry. We observe that, from Be10 clusters on, the deviation from CCSD(T) benchmarks is stable with respect to size, and fluctuating within 0.02 eV error bar for most examined functionals. This opens up the possibility of scaling the DFT binding energies for large Be clusters using CCSD(T) benchmark values for smaller clusters. We also tried to find analogies between the performance of DFT functionals for Be clusters and for the valence-isoelectronic Mg clusters investigated recently in Truhlar's group. We conclude that it is difficult to find DFT functionals that perform reasonably well for both beryllium and magnesium clusters. Out of 12 functionals examined, only the M06-2X functional gives reasonably accurate and balanced binding energies for both Be and Mg clusters.

The method of parallel-hierarchical transformation for rapid recognition of dynamic images using GPGPU technology

NASA Astrophysics Data System (ADS)

Timchenko, Leonid; Yarovyi, Andrii; Kokriatskaya, Nataliya; Nakonechna, Svitlana; Abramenko, Ludmila; Ławicki, Tomasz; Popiel, Piotr; Yesmakhanova, Laura

2016-09-01

The paper presents a method of parallel-hierarchical transformations for rapid recognition of dynamic images using GPU technology. Direct parallel-hierarchical transformations based on cluster CPU-and GPU-oriented hardware platform. Mathematic models of training of the parallel hierarchical (PH) network for the transformation are developed, as well as a training method of the PH network for recognition of dynamic images. This research is most topical for problems on organizing high-performance computations of super large arrays of information designed to implement multi-stage sensing and processing as well as compaction and recognition of data in the informational structures and computer devices. This method has such advantages as high performance through the use of recent advances in parallelization, possibility to work with images of ultra dimension, ease of scaling in case of changing the number of nodes in the cluster, auto scan of local network to detect compute nodes.

Patterns of time use among low-income urban minority adolescents and associations with academic outcomes and problem behaviors.

PubMed

Wolf, Sharon; Aber, J Lawrence; Morris, Pamela A

2015-06-01

Time budgets represent key opportunities for developmental support and contribute to an understanding of achievement gaps and adjustment across populations of youth. This study assessed the connection between out-of-school time use patterns and academic performance outcomes, academic motivations and goals, and problem behaviors for 504 low-income urban African American and Latino adolescents (54% female; M = 16.6 years). Time use patterns were measured across eight activity types using cluster analysis. Four groups of adolescents were identified, based on their different profiles of time use: (1) Academic: those with most time in academic activities; (2) Social: those with most time in social activities; (3) Maintenance/work: those with most time in maintenance and work activities; and (4) TV/computer: those with most time in TV or computer activities. Time use patterns were meaningfully associated with variation in outcomes in this population. Adolescents in the Academic cluster had the highest levels of adjustment across all domains; adolescents in the Social cluster had the lowest academic performance and highest problem behaviors; and adolescents in the TV/computer cluster had the lowest levels of intrinsic motivation. Females were more likely to be in the Academic cluster, and less likely to be in the other three clusters compared to males. No differences by race or gender were found in assessing the relationship between time use and outcomes. The study's results indicate that time use patterns are meaningfully associated with within-group variation in adjustment for low-income minority adolescents, and that shared contexts may shape time use more than individual differences in race/ethnicity for this population.

Improving real-time efficiency of case-based reasoning for medical diagnosis.

PubMed

Park, Yoon-Joo

2014-01-01

Conventional case-based reasoning (CBR) does not perform efficiently for high volume dataset because of case-retrieval time. Some previous researches overcome this problem by clustering a case-base into several small groups, and retrieve neighbors within a corresponding group to a target case. However, this approach generally produces less accurate predictive performances than the conventional CBR. This paper suggests a new case-based reasoning method called the Clustering-Merging CBR (CM-CBR) which produces similar level of predictive performances than the conventional CBR with spending significantly less computational cost.

Strategy Generalization across Orientation Tasks: Testing a Computational Cognitive Model

DTIC Science & Technology

2008-07-01

arranged in groups ( clusters ). The space, itself, was divided into four quadrants, which had 1, 2, 3, and 4 objects, respectively. The arrangement of... clusters , of objects play an important role in the model’s performance, by providing some context for narrowing the search for the target to a portion of the...model uses a hierarchical approach to accomplish this. First, the model identifies a group or cluster of objects that contains the target. The number of

Effective scheme for partitioning covalent bonds in density-functional embedding theory: From molecules to extended covalent systems.

PubMed

Huang, Chen; Muñoz-García, Ana Belén; Pavone, Michele

2016-12-28

Density-functional embedding theory provides a general way to perform multi-physics quantum mechanics simulations of large-scale materials by dividing the total system's electron density into a cluster's density and its environment's density. It is then possible to compute the accurate local electronic structures and energetics of the embedded cluster with high-level methods, meanwhile retaining a low-level description of the environment. The prerequisite step in the density-functional embedding theory is the cluster definition. In covalent systems, cutting across the covalent bonds that connect the cluster and its environment leads to dangling bonds (unpaired electrons). These represent a major obstacle for the application of density-functional embedding theory to study extended covalent systems. In this work, we developed a simple scheme to define the cluster in covalent systems. Instead of cutting covalent bonds, we directly split the boundary atoms for maintaining the valency of the cluster. With this new covalent embedding scheme, we compute the dehydrogenation energies of several different molecules, as well as the binding energy of a cobalt atom on graphene. Well localized cluster densities are observed, which can facilitate the use of localized basis sets in high-level calculations. The results are found to converge faster with the embedding method than the other multi-physics approach ONIOM. This work paves the way to perform the density-functional embedding simulations of heterogeneous systems in which different types of chemical bonds are present.

Comparison of Monte Carlo simulated and measured performance parameters of miniPET scanner

NASA Astrophysics Data System (ADS)

Kis, S. A.; Emri, M.; Opposits, G.; Bükki, T.; Valastyán, I.; Hegyesi, Gy.; Imrek, J.; Kalinka, G.; Molnár, J.; Novák, D.; Végh, J.; Kerek, A.; Trón, L.; Balkay, L.

2007-02-01

In vivo imaging of small laboratory animals is a valuable tool in the development of new drugs. For this purpose, miniPET, an easy to scale modular small animal PET camera has been developed at our institutes. The system has four modules, which makes it possible to rotate the whole detector system around the axis of the field of view. Data collection and image reconstruction are performed using a data acquisition (DAQ) module with Ethernet communication facility and a computer cluster of commercial PCs. Performance tests were carried out to determine system parameters, such as energy resolution, sensitivity and noise equivalent count rate. A modified GEANT4-based GATE Monte Carlo software package was used to simulate PET data analogous to those of the performance measurements. GATE was run on a Linux cluster of 10 processors (64 bit, Xeon with 3.0 GHz) and controlled by a SUN grid engine. The application of this special computer cluster reduced the time necessary for the simulations by an order of magnitude. The simulated energy spectra, maximum rate of true coincidences and sensitivity of the camera were in good agreement with the measured parameters.

Federated data storage system prototype for LHC experiments and data intensive science

NASA Astrophysics Data System (ADS)

Kiryanov, A.; Klimentov, A.; Krasnopevtsev, D.; Ryabinkin, E.; Zarochentsev, A.

2017-10-01

Rapid increase of data volume from the experiments running at the Large Hadron Collider (LHC) prompted physics computing community to evaluate new data handling and processing solutions. Russian grid sites and universities’ clusters scattered over a large area aim at the task of uniting their resources for future productive work, at the same time giving an opportunity to support large physics collaborations. In our project we address the fundamental problem of designing a computing architecture to integrate distributed storage resources for LHC experiments and other data-intensive science applications and to provide access to data from heterogeneous computing facilities. Studies include development and implementation of federated data storage prototype for Worldwide LHC Computing Grid (WLCG) centres of different levels and University clusters within one National Cloud. The prototype is based on computing resources located in Moscow, Dubna, Saint Petersburg, Gatchina and Geneva. This project intends to implement a federated distributed storage for all kind of operations such as read/write/transfer and access via WAN from Grid centres, university clusters, supercomputers, academic and commercial clouds. The efficiency and performance of the system are demonstrated using synthetic and experiment-specific tests including real data processing and analysis workflows from ATLAS and ALICE experiments, as well as compute-intensive bioinformatics applications (PALEOMIX) running on supercomputers. We present topology and architecture of the designed system, report performance and statistics for different access patterns and show how federated data storage can be used efficiently by physicists and biologists. We also describe how sharing data on a widely distributed storage system can lead to a new computing model and reformations of computing style, for instance how bioinformatics program running on supercomputers can read/write data from the federated storage.

Relativistic coupled cluster theory based on the no-pair Dirac-Coulomb-Breit Hamiltonian: Relativistic pair correlation energies of the Xe atom

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

Eliav, E.; Kaldor, U.; Ishikawa, Y.

1994-12-31

Relativistic pair correlation energies of Xe were computed by employing a recently developed relativistic coupled cluster theory based on the no-pair Dirac-Coulomb-Breit Hamiltonian. The matrix Dirac-Fock-Breit SCF and relativistic coupled cluster calculations were performed by means of expansion in basis sets of well-tempered Gaussian spinors. A detailed study of the pair correlation energies in Xe is performed, in order to investigate the effects of the low-frequency Breit interaction on the correlation energies of Xe. Nonadditivity of correlation and relativistic (particularly Breit) effects is discussed.

HRLSim: a high performance spiking neural network simulator for GPGPU clusters.

PubMed

Minkovich, Kirill; Thibeault, Corey M; O'Brien, Michael John; Nogin, Aleksey; Cho, Youngkwan; Srinivasa, Narayan

2014-02-01

Modeling of large-scale spiking neural models is an important tool in the quest to understand brain function and subsequently create real-world applications. This paper describes a spiking neural network simulator environment called HRL Spiking Simulator (HRLSim). This simulator is suitable for implementation on a cluster of general purpose graphical processing units (GPGPUs). Novel aspects of HRLSim are described and an analysis of its performance is provided for various configurations of the cluster. With the advent of inexpensive GPGPU cards and compute power, HRLSim offers an affordable and scalable tool for design, real-time simulation, and analysis of large-scale spiking neural networks.

An improved method to detect correct protein folds using partial clustering.

PubMed

Zhou, Jianjun; Wishart, David S

2013-01-16

Structure-based clustering is commonly used to identify correct protein folds among candidate folds (also called decoys) generated by protein structure prediction programs. However, traditional clustering methods exhibit a poor runtime performance on large decoy sets. We hypothesized that a more efficient "partial" clustering approach in combination with an improved scoring scheme could significantly improve both the speed and performance of existing candidate selection methods. We propose a new scheme that performs rapid but incomplete clustering on protein decoys. Our method detects structurally similar decoys (measured using either C(α) RMSD or GDT-TS score) and extracts representatives from them without assigning every decoy to a cluster. We integrated our new clustering strategy with several different scoring functions to assess both the performance and speed in identifying correct or near-correct folds. Experimental results on 35 Rosetta decoy sets and 40 I-TASSER decoy sets show that our method can improve the correct fold detection rate as assessed by two different quality criteria. This improvement is significantly better than two recently published clustering methods, Durandal and Calibur-lite. Speed and efficiency testing shows that our method can handle much larger decoy sets and is up to 22 times faster than Durandal and Calibur-lite. The new method, named HS-Forest, avoids the computationally expensive task of clustering every decoy, yet still allows superior correct-fold selection. Its improved speed, efficiency and decoy-selection performance should enable structure prediction researchers to work with larger decoy sets and significantly improve their ab initio structure prediction performance.

An improved method to detect correct protein folds using partial clustering

PubMed Central

2013-01-01

Background Structure-based clustering is commonly used to identify correct protein folds among candidate folds (also called decoys) generated by protein structure prediction programs. However, traditional clustering methods exhibit a poor runtime performance on large decoy sets. We hypothesized that a more efficient “partial“ clustering approach in combination with an improved scoring scheme could significantly improve both the speed and performance of existing candidate selection methods. Results We propose a new scheme that performs rapid but incomplete clustering on protein decoys. Our method detects structurally similar decoys (measured using either Cα RMSD or GDT-TS score) and extracts representatives from them without assigning every decoy to a cluster. We integrated our new clustering strategy with several different scoring functions to assess both the performance and speed in identifying correct or near-correct folds. Experimental results on 35 Rosetta decoy sets and 40 I-TASSER decoy sets show that our method can improve the correct fold detection rate as assessed by two different quality criteria. This improvement is significantly better than two recently published clustering methods, Durandal and Calibur-lite. Speed and efficiency testing shows that our method can handle much larger decoy sets and is up to 22 times faster than Durandal and Calibur-lite. Conclusions The new method, named HS-Forest, avoids the computationally expensive task of clustering every decoy, yet still allows superior correct-fold selection. Its improved speed, efficiency and decoy-selection performance should enable structure prediction researchers to work with larger decoy sets and significantly improve their ab initio structure prediction performance. PMID:23323835

Comparing the OpenMP, MPI, and Hybrid Programming Paradigm on an SMP Cluster

NASA Technical Reports Server (NTRS)

Jost, Gabriele; Jin, Haoqiang; anMey, Dieter; Hatay, Ferhat F.

2003-01-01

With the advent of parallel hardware and software technologies users are faced with the challenge to choose a programming paradigm best suited for the underlying computer architecture. With the current trend in parallel computer architectures towards clusters of shared memory symmetric multi-processors (SMP), parallel programming techniques have evolved to support parallelism beyond a single level. Which programming paradigm is the best will depend on the nature of the given problem, the hardware architecture, and the available software. In this study we will compare different programming paradigms for the parallelization of a selected benchmark application on a cluster of SMP nodes. We compare the timings of different implementations of the same CFD benchmark application employing the same numerical algorithm on a cluster of Sun Fire SMP nodes. The rest of the paper is structured as follows: In section 2 we briefly discuss the programming models under consideration. We describe our compute platform in section 3. The different implementations of our benchmark code are described in section 4 and the performance results are presented in section 5. We conclude our study in section 6.

StrAuto: automation and parallelization of STRUCTURE analysis.

PubMed

Chhatre, Vikram E; Emerson, Kevin J

2017-03-24

Population structure inference using the software STRUCTURE has become an integral part of population genetic studies covering a broad spectrum of taxa including humans. The ever-expanding size of genetic data sets poses computational challenges for this analysis. Although at least one tool currently implements parallel computing to reduce computational overload of this analysis, it does not fully automate the use of replicate STRUCTURE analysis runs required for downstream inference of optimal K. There is pressing need for a tool that can deploy population structure analysis on high performance computing clusters. We present an updated version of the popular Python program StrAuto, to streamline population structure analysis using parallel computing. StrAuto implements a pipeline that combines STRUCTURE analysis with the Evanno Δ K analysis and visualization of results using STRUCTURE HARVESTER. Using benchmarking tests, we demonstrate that StrAuto significantly reduces the computational time needed to perform iterative STRUCTURE analysis by distributing runs over two or more processors. StrAuto is the first tool to integrate STRUCTURE analysis with post-processing using a pipeline approach in addition to implementing parallel computation - a set up ideal for deployment on computing clusters. StrAuto is distributed under the GNU GPL (General Public License) and available to download from http://strauto.popgen.org .

Image Segmentation Method Using Fuzzy C Mean Clustering Based on Multi-Objective Optimization

NASA Astrophysics Data System (ADS)

Chen, Jinlin; Yang, Chunzhi; Xu, Guangkui; Ning, Li

2018-04-01

Image segmentation is not only one of the hottest topics in digital image processing, but also an important part of computer vision applications. As one kind of image segmentation algorithms, fuzzy C-means clustering is an effective and concise segmentation algorithm. However, the drawback of FCM is that it is sensitive to image noise. To solve the problem, this paper designs a novel fuzzy C-mean clustering algorithm based on multi-objective optimization. We add a parameter λ to the fuzzy distance measurement formula to improve the multi-objective optimization. The parameter λ can adjust the weights of the pixel local information. In the algorithm, the local correlation of neighboring pixels is added to the improved multi-objective mathematical model to optimize the clustering cent. Two different experimental results show that the novel fuzzy C-means approach has an efficient performance and computational time while segmenting images by different type of noises.

Removal of impulse noise clusters from color images with local order statistics

NASA Astrophysics Data System (ADS)

Ruchay, Alexey; Kober, Vitaly

2017-09-01

This paper proposes a novel algorithm for restoring images corrupted with clusters of impulse noise. The noise clusters often occur when the probability of impulse noise is very high. The proposed noise removal algorithm consists of detection of bulky impulse noise in three color channels with local order statistics followed by removal of the detected clusters by means of vector median filtering. With the help of computer simulation we show that the proposed algorithm is able to effectively remove clustered impulse noise. The performance of the proposed algorithm is compared in terms of image restoration metrics with that of common successful algorithms.

Double Cluster Heads Model for Secure and Accurate Data Fusion in Wireless Sensor Networks

PubMed Central

Fu, Jun-Song; Liu, Yun

2015-01-01

Secure and accurate data fusion is an important issue in wireless sensor networks (WSNs) and has been extensively researched in the literature. In this paper, by combining clustering techniques, reputation and trust systems, and data fusion algorithms, we propose a novel cluster-based data fusion model called Double Cluster Heads Model (DCHM) for secure and accurate data fusion in WSNs. Different from traditional clustering models in WSNs, two cluster heads are selected after clustering for each cluster based on the reputation and trust system and they perform data fusion independently of each other. Then, the results are sent to the base station where the dissimilarity coefficient is computed. If the dissimilarity coefficient of the two data fusion results exceeds the threshold preset by the users, the cluster heads will be added to blacklist, and the cluster heads must be reelected by the sensor nodes in a cluster. Meanwhile, feedback is sent from the base station to the reputation and trust system, which can help us to identify and delete the compromised sensor nodes in time. Through a series of extensive simulations, we found that the DCHM performed very well in data fusion security and accuracy. PMID:25608211

a Linux PC Cluster for Lattice QCD with Exact Chiral Symmetry

NASA Astrophysics Data System (ADS)

Chiu, Ting-Wai; Hsieh, Tung-Han; Huang, Chao-Hsi; Huang, Tsung-Ren

A computational system for lattice QCD with overlap Dirac quarks is described. The platform is a home-made Linux PC cluster, built with off-the-shelf components. At present the system constitutes of 64 nodes, with each node consisting of one Pentium 4 processor (1.6/2.0/2.5 GHz), one Gbyte of PC800/1066 RDRAM, one 40/80/120 Gbyte hard disk, and a network card. The computationally intensive parts of our program are written in SSE2 codes. The speed of our system is estimated to be 70 Gflops, and its price/performance ratio is better than $1.0/Mflops for 64-bit (double precision) computations in quenched QCD. We discuss how to optimize its hardware and software for computing propagators of overlap Dirac quarks.

Multiscale Computer Simulation of Failure in Aerogels

NASA Technical Reports Server (NTRS)

Good, Brian S.

2008-01-01

Aerogels have been of interest to the aerospace community primarily for their thermal properties, notably their low thermal conductivities. While such gels are typically fragile, recent advances in the application of conformal polymer layers to these gels has made them potentially useful as lightweight structural materials as well. We have previously performed computer simulations of aerogel thermal conductivity and tensile and compressive failure, with results that are in qualitative, and sometimes quantitative, agreement with experiment. However, recent experiments in our laboratory suggest that gels having similar densities may exhibit substantially different properties. In this work, we extend our original diffusion limited cluster aggregation (DLCA) model for gel structure to incorporate additional variation in DLCA simulation parameters, with the aim of producing DLCA clusters of similar densities that nevertheless have different fractal dimension and secondary particle coordination. We perform particle statics simulations of gel strain on these clusters, and consider the effects of differing DLCA simulation conditions, and the resultant differences in fractal dimension and coordination, on gel strain properties.

Performance Analysis Tool for HPC and Big Data Applications on Scientific Clusters

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

Yoo, Wucherl; Koo, Michelle; Cao, Yu

Big data is prevalent in HPC computing. Many HPC projects rely on complex workflows to analyze terabytes or petabytes of data. These workflows often require running over thousands of CPU cores and performing simultaneous data accesses, data movements, and computation. It is challenging to analyze the performance involving terabytes or petabytes of workflow data or measurement data of the executions, from complex workflows over a large number of nodes and multiple parallel task executions. To help identify performance bottlenecks or debug the performance issues in large-scale scientific applications and scientific clusters, we have developed a performance analysis framework, using state-ofthe-more » art open-source big data processing tools. Our tool can ingest system logs and application performance measurements to extract key performance features, and apply the most sophisticated statistical tools and data mining methods on the performance data. It utilizes an efficient data processing engine to allow users to interactively analyze a large amount of different types of logs and measurements. To illustrate the functionality of the big data analysis framework, we conduct case studies on the workflows from an astronomy project known as the Palomar Transient Factory (PTF) and the job logs from the genome analysis scientific cluster. Our study processed many terabytes of system logs and application performance measurements collected on the HPC systems at NERSC. The implementation of our tool is generic enough to be used for analyzing the performance of other HPC systems and Big Data workows.« less

Excess electrons in methanol clusters: Beyond the one-electron picture

NASA Astrophysics Data System (ADS)

Pohl, Gábor; Mones, Letif; Turi, László

2016-10-01

We performed a series of comparative quantum chemical calculations on various size negatively charged methanol clusters, ("separators=" CH 3 OH ) n - . The clusters are examined in their optimized geometries (n = 2-4), and in geometries taken from mixed quantum-classical molecular dynamics simulations at finite temperature (n = 2-128). These latter structures model potential electron binding sites in methanol clusters and in bulk methanol. In particular, we compute the vertical detachment energy (VDE) of an excess electron from increasing size methanol cluster anions using quantum chemical computations at various levels of theory including a one-electron pseudopotential model, several density functional theory (DFT) based methods, MP2 and coupled-cluster CCSD(T) calculations. The results suggest that at least four methanol molecules are needed to bind an excess electron on a hydrogen bonded methanol chain in a dipole bound state. Larger methanol clusters are able to form stronger interactions with an excess electron. The two simulated excess electron binding motifs in methanol clusters, interior and surface states, correlate well with distinct, experimentally found VDE tendencies with size. Interior states in a solvent cavity are stabilized significantly stronger than electron states on cluster surfaces. Although we find that all the examined quantum chemistry methods more or less overestimate the strength of the experimental excess electron stabilization, MP2, LC-BLYP, and BHandHLYP methods with diffuse basis sets provide a significantly better estimate of the VDE than traditional DFT methods (BLYP, B3LYP, X3LYP, PBE0). A comparison to the better performing many electron methods indicates that the examined one-electron pseudopotential can be reasonably used in simulations for systems of larger size.

Excess electrons in methanol clusters: Beyond the one-electron picture.

PubMed

Pohl, Gábor; Mones, Letif; Turi, László

2016-10-28

We performed a series of comparative quantum chemical calculations on various size negatively charged methanol clusters, CH 3 OH n - . The clusters are examined in their optimized geometries (n = 2-4), and in geometries taken from mixed quantum-classical molecular dynamics simulations at finite temperature (n = 2-128). These latter structures model potential electron binding sites in methanol clusters and in bulk methanol. In particular, we compute the vertical detachment energy (VDE) of an excess electron from increasing size methanol cluster anions using quantum chemical computations at various levels of theory including a one-electron pseudopotential model, several density functional theory (DFT) based methods, MP2 and coupled-cluster CCSD(T) calculations. The results suggest that at least four methanol molecules are needed to bind an excess electron on a hydrogen bonded methanol chain in a dipole bound state. Larger methanol clusters are able to form stronger interactions with an excess electron. The two simulated excess electron binding motifs in methanol clusters, interior and surface states, correlate well with distinct, experimentally found VDE tendencies with size. Interior states in a solvent cavity are stabilized significantly stronger than electron states on cluster surfaces. Although we find that all the examined quantum chemistry methods more or less overestimate the strength of the experimental excess electron stabilization, MP2, LC-BLYP, and BHandHLYP methods with diffuse basis sets provide a significantly better estimate of the VDE than traditional DFT methods (BLYP, B3LYP, X3LYP, PBE0). A comparison to the better performing many electron methods indicates that the examined one-electron pseudopotential can be reasonably used in simulations for systems of larger size.

A comparison of heuristic and model-based clustering methods for dietary pattern analysis.

PubMed

Greve, Benjamin; Pigeot, Iris; Huybrechts, Inge; Pala, Valeria; Börnhorst, Claudia

2016-02-01

Cluster analysis is widely applied to identify dietary patterns. A new method based on Gaussian mixture models (GMM) seems to be more flexible compared with the commonly applied k-means and Ward's method. In the present paper, these clustering approaches are compared to find the most appropriate one for clustering dietary data. The clustering methods were applied to simulated data sets with different cluster structures to compare their performance knowing the true cluster membership of observations. Furthermore, the three methods were applied to FFQ data assessed in 1791 children participating in the IDEFICS (Identification and Prevention of Dietary- and Lifestyle-Induced Health Effects in Children and Infants) Study to explore their performance in practice. The GMM outperformed the other methods in the simulation study in 72 % up to 100 % of cases, depending on the simulated cluster structure. Comparing the computationally less complex k-means and Ward's methods, the performance of k-means was better in 64-100 % of cases. Applied to real data, all methods identified three similar dietary patterns which may be roughly characterized as a 'non-processed' cluster with a high consumption of fruits, vegetables and wholemeal bread, a 'balanced' cluster with only slight preferences of single foods and a 'junk food' cluster. The simulation study suggests that clustering via GMM should be preferred due to its higher flexibility regarding cluster volume, shape and orientation. The k-means seems to be a good alternative, being easier to use while giving similar results when applied to real data.

Taming Pipelines, Users, and High Performance Computing with Rector

NASA Astrophysics Data System (ADS)

Estes, N. M.; Bowley, K. S.; Paris, K. N.; Silva, V. H.; Robinson, M. S.

2018-04-01

Rector is a high-performance job management system created by the LROC SOC team to enable processing of thousands of observations and ancillary data products as well as ad-hoc user jobs across a 634 CPU core processing cluster.

Multi-agent grid system Agent-GRID with dynamic load balancing of cluster nodes

NASA Astrophysics Data System (ADS)

Satymbekov, M. N.; Pak, I. T.; Naizabayeva, L.; Nurzhanov, Ch. A.

2017-12-01

In this study the work presents the system designed for automated load balancing of the contributor by analysing the load of compute nodes and the subsequent migration of virtual machines from loaded nodes to less loaded ones. This system increases the performance of cluster nodes and helps in the timely processing of data. A grid system balances the work of cluster nodes the relevance of the system is the award of multi-agent balancing for the solution of such problems.

Fluid{Structure Interaction Modeling of Modified-Porosity Parachutes and Parachute Clusters

NASA Astrophysics Data System (ADS)

Boben, Joseph J.

To increase aerodynamic performance, the geometric porosity of a ringsail spacecraft parachute canopy is sometimes increased, beyond the "rings" and "sails" with hundreds of "ring gaps" and "sail slits." This creates extra computational challenges for fluid-structure interaction (FSI) modeling of clusters of such parachutes, beyond those created by the lightness of the canopy structure, geometric complexities of hundreds of gaps and slits, and the contact between the parachutes of the cluster. In FSI computation of parachutes with such "modified geometric porosity," the ow through the "windows" created by the removal of the panels and the wider gaps created by the removal of the sails cannot be accurately modeled with the Homogenized Modeling of Geometric Porosity (HMGP), which was introduced to deal with the hundreds of gaps and slits. The ow needs to be actually resolved. All these computational challenges need to be addressed simultaneously in FSI modeling of clusters of spacecraft parachutes with modified geometric porosity. The core numerical technology is the Stabilized Space-Time FSI (SSTFSI) technique, and the contact between the parachutes is handled with the Surface-Edge-Node Contact Tracking (SENCT) technique. In the computations reported here, in addition to the SSTFSI and SENCT techniques and HMGP, we use the special techniques we have developed for removing the numerical spinning component of the parachute motion and for restoring the mesh integrity without a remesh. We present results for 2- and 3-parachute clusters with two different payload models. We also present the FSI computations we carried out for a single, subscale modified-porosity parachute.

[Earth Science Technology Office's Computational Technologies Project

NASA Technical Reports Server (NTRS)

Fischer, James (Technical Monitor); Merkey, Phillip

2005-01-01

This grant supported the effort to characterize the problem domain of the Earth Science Technology Office's Computational Technologies Project, to engage the Beowulf Cluster Computing Community as well as the High Performance Computing Research Community so that we can predict the applicability of said technologies to the scientific community represented by the CT project and formulate long term strategies to provide the computational resources necessary to attain the anticipated scientific objectives of the CT project. Specifically, the goal of the evaluation effort is to use the information gathered over the course of the Round-3 investigations to quantify the trends in scientific expectations, the algorithmic requirements and capabilities of high-performance computers to satisfy this anticipated need.

Visualization of Unsteady Computational Fluid Dynamics

NASA Technical Reports Server (NTRS)

Haimes, Robert

1997-01-01

The current compute environment that most researchers are using for the calculation of 3D unsteady Computational Fluid Dynamic (CFD) results is a super-computer class machine. The Massively Parallel Processors (MPP's) such as the 160 node IBM SP2 at NAS and clusters of workstations acting as a single MPP (like NAS's SGI Power-Challenge array and the J90 cluster) provide the required computation bandwidth for CFD calculations of transient problems. If we follow the traditional computational analysis steps for CFD (and we wish to construct an interactive visualizer) we need to be aware of the following: (1) Disk space requirements. A single snap-shot must contain at least the values (primitive variables) stored at the appropriate locations within the mesh. For most simple 3D Euler solvers that means 5 floating point words. Navier-Stokes solutions with turbulence models may contain 7 state-variables. (2) Disk speed vs. Computational speeds. The time required to read the complete solution of a saved time frame from disk is now longer than the compute time for a set number of iterations from an explicit solver. Depending, on the hardware and solver an iteration of an implicit code may also take less time than reading the solution from disk. If one examines the performance improvements in the last decade or two, it is easy to see that depending on disk performance (vs. CPU improvement) may not be the best method for enhancing interactivity. (3) Cluster and Parallel Machine I/O problems. Disk access time is much worse within current parallel machines and cluster of workstations that are acting in concert to solve a single problem. In this case we are not trying to read the volume of data, but are running the solver and the solver outputs the solution. These traditional network interfaces must be used for the file system. (4) Numerics of particle traces. Most visualization tools can work upon a single snap shot of the data but some visualization tools for transient problems require dealing with time.

Implementation of MPEG-2 encoder to multiprocessor system using multiple MVPs (TMS320C80)

NASA Astrophysics Data System (ADS)

Kim, HyungSun; Boo, Kenny; Chung, SeokWoo; Choi, Geon Y.; Lee, YongJin; Jeon, JaeHo; Park, Hyun Wook

1997-05-01

This paper presents the efficient algorithm mapping for the real-time MPEG-2 encoding on the KAIST image computing system (KICS), which has a parallel architecture using five multimedia video processors (MVPs). The MVP is a general purpose digital signal processor (DSP) of Texas Instrument. It combines one floating-point processor and four fixed- point DSPs on a single chip. The KICS uses the MVP as a primary processing element (PE). Two PEs form a cluster, and there are two processing clusters in the KICS. Real-time MPEG-2 encoder is implemented through the spatial and the functional partitioning strategies. Encoding process of spatially partitioned half of the video input frame is assigned to ne processing cluster. Two PEs perform the functionally partitioned MPEG-2 encoding tasks in the pipelined operation mode. One PE of a cluster carries out the transform coding part and the other performs the predictive coding part of the MPEG-2 encoding algorithm. One MVP among five MVPs is used for system control and interface with host computer. This paper introduces an implementation of the MPEG-2 algorithm with a parallel processing architecture.

Self-consistent clustering analysis: an efficient multiscale scheme for inelastic heterogeneous materials

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

Liu, Z.; Bessa, M. A.; Liu, W.K.

A predictive computational theory is shown for modeling complex, hierarchical materials ranging from metal alloys to polymer nanocomposites. The theory can capture complex mechanisms such as plasticity and failure that span across multiple length scales. This general multiscale material modeling theory relies on sound principles of mathematics and mechanics, and a cutting-edge reduced order modeling method named self-consistent clustering analysis (SCA) [Zeliang Liu, M.A. Bessa, Wing Kam Liu, “Self-consistent clustering analysis: An efficient multi-scale scheme for inelastic heterogeneous materials,” Comput. Methods Appl. Mech. Engrg. 306 (2016) 319–341]. SCA reduces by several orders of magnitude the computational cost of micromechanical andmore » concurrent multiscale simulations, while retaining the microstructure information. This remarkable increase in efficiency is achieved with a data-driven clustering method. Computationally expensive operations are performed in the so-called offline stage, where degrees of freedom (DOFs) are agglomerated into clusters. The interaction tensor of these clusters is computed. In the online or predictive stage, the Lippmann-Schwinger integral equation is solved cluster-wise using a self-consistent scheme to ensure solution accuracy and avoid path dependence. To construct a concurrent multiscale model, this scheme is applied at each material point in a macroscale structure, replacing a conventional constitutive model with the average response computed from the microscale model using just the SCA online stage. A regularized damage theory is incorporated in the microscale that avoids the mesh and RVE size dependence that commonly plagues microscale damage calculations. The SCA method is illustrated with two cases: a carbon fiber reinforced polymer (CFRP) structure with the concurrent multiscale model and an application to fatigue prediction for additively manufactured metals. For the CFRP problem, a speed up estimated to be about 43,000 is achieved by using the SCA method, as opposed to FE2, enabling the solution of an otherwise computationally intractable problem. The second example uses a crystal plasticity constitutive law and computes the fatigue potency of extrinsic microscale features such as voids. This shows that local stress and strain are capture sufficiently well by SCA. This model has been incorporated in a process-structure-properties prediction framework for process design in additive manufacturing.« less

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/.

Quantum computational universality of the Cai-Miyake-Dür-Briegel two-dimensional quantum state from Affleck-Kennedy-Lieb-Tasaki quasichains

NASA Astrophysics Data System (ADS)

Wei, Tzu-Chieh; Raussendorf, Robert; Kwek, Leong Chuan

2011-10-01

Universal quantum computation can be achieved by simply performing single-qubit measurements on a highly entangled resource state, such as cluster states. Cai, Miyake, Dür, and Briegel recently constructed a ground state of a two-dimensional quantum magnet by combining multiple Affleck-Kennedy-Lieb-Tasaki quasichains of mixed spin-3/2 and spin-1/2 entities and by mapping pairs of neighboring spin-1/2 particles to individual spin-3/2 particles [Phys. Rev. APLRAAN1050-294710.1103/PhysRevA.82.052309 82, 052309 (2010)]. They showed that this state enables universal quantum computation by single-spin measurements. Here, we give an alternative understanding of how this state gives rise to universal measurement-based quantum computation: by local operations, each quasichain can be converted to a one-dimensional cluster state and entangling gates between two neighboring logical qubits can be implemented by single-spin measurements. We further argue that a two-dimensional cluster state can be distilled from the Cai-Miyake-Dür-Briegel state.

RosettaAntibodyDesign (RAbD): A general framework for computational antibody design

PubMed Central

Adolf-Bryfogle, Jared; Kalyuzhniy, Oleks; Kubitz, Michael; Hu, Xiaozhen; Adachi, Yumiko; Schief, William R.

2018-01-01

A structural-bioinformatics-based computational methodology and framework have been developed for the design of antibodies to targets of interest. RosettaAntibodyDesign (RAbD) samples the diverse sequence, structure, and binding space of an antibody to an antigen in highly customizable protocols for the design of antibodies in a broad range of applications. The program samples antibody sequences and structures by grafting structures from a widely accepted set of the canonical clusters of CDRs (North et al., J. Mol. Biol., 406:228–256, 2011). It then performs sequence design according to amino acid sequence profiles of each cluster, and samples CDR backbones using a flexible-backbone design protocol incorporating cluster-based CDR constraints. Starting from an existing experimental or computationally modeled antigen-antibody structure, RAbD can be used to redesign a single CDR or multiple CDRs with loops of different length, conformation, and sequence. We rigorously benchmarked RAbD on a set of 60 diverse antibody–antigen complexes, using two design strategies—optimizing total Rosetta energy and optimizing interface energy alone. We utilized two novel metrics for measuring success in computational protein design. The design risk ratio (DRR) is equal to the frequency of recovery of native CDR lengths and clusters divided by the frequency of sampling of those features during the Monte Carlo design procedure. Ratios greater than 1.0 indicate that the design process is picking out the native more frequently than expected from their sampled rate. We achieved DRRs for the non-H3 CDRs of between 2.4 and 4.0. The antigen risk ratio (ARR) is the ratio of frequencies of the native amino acid types, CDR lengths, and clusters in the output decoys for simulations performed in the presence and absence of the antigen. For CDRs, we achieved cluster ARRs as high as 2.5 for L1 and 1.5 for H2. For sequence design simulations without CDR grafting, the overall recovery for the native amino acid types for residues that contact the antigen in the native structures was 72% in simulations performed in the presence of the antigen and 48% in simulations performed without the antigen, for an ARR of 1.5. For the non-contacting residues, the ARR was 1.08. This shows that the sequence profiles are able to maintain the amino acid types of these conserved, buried sites, while recovery of the exposed, contacting residues requires the presence of the antigen-antibody interface. We tested RAbD experimentally on both a lambda and kappa antibody–antigen complex, successfully improving their affinities 10 to 50 fold by replacing individual CDRs of the native antibody with new CDR lengths and clusters. PMID:29702641

RosettaAntibodyDesign (RAbD): A general framework for computational antibody design.

PubMed

Adolf-Bryfogle, Jared; Kalyuzhniy, Oleks; Kubitz, Michael; Weitzner, Brian D; Hu, Xiaozhen; Adachi, Yumiko; Schief, William R; Dunbrack, Roland L

2018-04-01

A structural-bioinformatics-based computational methodology and framework have been developed for the design of antibodies to targets of interest. RosettaAntibodyDesign (RAbD) samples the diverse sequence, structure, and binding space of an antibody to an antigen in highly customizable protocols for the design of antibodies in a broad range of applications. The program samples antibody sequences and structures by grafting structures from a widely accepted set of the canonical clusters of CDRs (North et al., J. Mol. Biol., 406:228-256, 2011). It then performs sequence design according to amino acid sequence profiles of each cluster, and samples CDR backbones using a flexible-backbone design protocol incorporating cluster-based CDR constraints. Starting from an existing experimental or computationally modeled antigen-antibody structure, RAbD can be used to redesign a single CDR or multiple CDRs with loops of different length, conformation, and sequence. We rigorously benchmarked RAbD on a set of 60 diverse antibody-antigen complexes, using two design strategies-optimizing total Rosetta energy and optimizing interface energy alone. We utilized two novel metrics for measuring success in computational protein design. The design risk ratio (DRR) is equal to the frequency of recovery of native CDR lengths and clusters divided by the frequency of sampling of those features during the Monte Carlo design procedure. Ratios greater than 1.0 indicate that the design process is picking out the native more frequently than expected from their sampled rate. We achieved DRRs for the non-H3 CDRs of between 2.4 and 4.0. The antigen risk ratio (ARR) is the ratio of frequencies of the native amino acid types, CDR lengths, and clusters in the output decoys for simulations performed in the presence and absence of the antigen. For CDRs, we achieved cluster ARRs as high as 2.5 for L1 and 1.5 for H2. For sequence design simulations without CDR grafting, the overall recovery for the native amino acid types for residues that contact the antigen in the native structures was 72% in simulations performed in the presence of the antigen and 48% in simulations performed without the antigen, for an ARR of 1.5. For the non-contacting residues, the ARR was 1.08. This shows that the sequence profiles are able to maintain the amino acid types of these conserved, buried sites, while recovery of the exposed, contacting residues requires the presence of the antigen-antibody interface. We tested RAbD experimentally on both a lambda and kappa antibody-antigen complex, successfully improving their affinities 10 to 50 fold by replacing individual CDRs of the native antibody with new CDR lengths and clusters.

A note on the kappa statistic for clustered dichotomous data.

PubMed

Zhou, Ming; Yang, Zhao

2014-06-30

The kappa statistic is widely used to assess the agreement between two raters. Motivated by a simulation-based cluster bootstrap method to calculate the variance of the kappa statistic for clustered physician-patients dichotomous data, we investigate its special correlation structure and develop a new simple and efficient data generation algorithm. For the clustered physician-patients dichotomous data, based on the delta method and its special covariance structure, we propose a semi-parametric variance estimator for the kappa statistic. An extensive Monte Carlo simulation study is performed to evaluate the performance of the new proposal and five existing methods with respect to the empirical coverage probability, root-mean-square error, and average width of the 95% confidence interval for the kappa statistic. The variance estimator ignoring the dependence within a cluster is generally inappropriate, and the variance estimators from the new proposal, bootstrap-based methods, and the sampling-based delta method perform reasonably well for at least a moderately large number of clusters (e.g., the number of clusters K ⩾50). The new proposal and sampling-based delta method provide convenient tools for efficient computations and non-simulation-based alternatives to the existing bootstrap-based methods. Moreover, the new proposal has acceptable performance even when the number of clusters is as small as K = 25. To illustrate the practical application of all the methods, one psychiatric research data and two simulated clustered physician-patients dichotomous data are analyzed. Copyright © 2014 John Wiley & Sons, Ltd.

Cluster-guided imaging-based CFD analysis of airflow and particle deposition in asthmatic human lungs

NASA Astrophysics Data System (ADS)

Choi, Jiwoong; Leblanc, Lawrence; Choi, Sanghun; Haghighi, Babak; Hoffman, Eric; Lin, Ching-Long

2017-11-01

The goal of this study is to assess inter-subject variability in delivery of orally inhaled drug products to small airways in asthmatic lungs. A recent multiscale imaging-based cluster analysis (MICA) of computed tomography (CT) lung images in an asthmatic cohort identified four clusters with statistically distinct structural and functional phenotypes associating with unique clinical biomarkers. Thus, we aimed to address inter-subject variability via inter-cluster variability. We selected a representative subject from each of the 4 asthma clusters as well as 1 male and 1 female healthy controls, and performed computational fluid and particle simulations on CT-based airway models of these subjects. The results from one severe and one non-severe asthmatic cluster subjects characterized by segmental airway constriction had increased particle deposition efficiency, as compared with the other two cluster subjects (one non-severe and one severe asthmatics) without airway constriction. Constriction-induced jets impinging on distal bifurcations led to excessive particle deposition. The results emphasize the impact of airway constriction on regional particle deposition rather than disease severity, demonstrating the potential of using cluster membership to tailor drug delivery. NIH Grants U01HL114494 and S10-RR022421, and FDA Grant U01FD005837. XSEDE.

A comparison of queueing, cluster and distributed computing systems

NASA Technical Reports Server (NTRS)

Kaplan, Joseph A.; Nelson, Michael L.

1993-01-01

Using workstation clusters for distributed computing has become popular with the proliferation of inexpensive, powerful workstations. Workstation clusters offer both a cost effective alternative to batch processing and an easy entry into parallel computing. However, a number of workstations on a network does not constitute a cluster. Cluster management software is necessary to harness the collective computing power. A variety of cluster management and queuing systems are compared: Distributed Queueing Systems (DQS), Condor, Load Leveler, Load Balancer, Load Sharing Facility (LSF - formerly Utopia), Distributed Job Manager (DJM), Computing in Distributed Networked Environments (CODINE), and NQS/Exec. The systems differ in their design philosophy and implementation. Based on published reports on the different systems and conversations with the system's developers and vendors, a comparison of the systems are made on the integral issues of clustered computing.

Harry Sorensen | NREL

Science.gov Websites

NREL's high performance computing cluster. Education B.S., engineering, Colorado School of Mines and platforms. His experience with writing engineering applications in Python is being utilized for

Understanding I/O workload characteristics of a Peta-scale storage system

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

Kim, Youngjae; Gunasekaran, Raghul

2015-01-01

Understanding workload characteristics is critical for optimizing and improving the performance of current systems and software, and architecting new storage systems based on observed workload patterns. In this paper, we characterize the I/O workloads of scientific applications of one of the world s fastest high performance computing (HPC) storage cluster, Spider, at the Oak Ridge Leadership Computing Facility (OLCF). OLCF flagship petascale simulation platform, Titan, and other large HPC clusters, in total over 250 thousands compute cores, depend on Spider for their I/O needs. We characterize the system utilization, the demands of reads and writes, idle time, storage space utilization,more » and the distribution of read requests to write requests for the Peta-scale Storage Systems. From this study, we develop synthesized workloads, and we show that the read and write I/O bandwidth usage as well as the inter-arrival time of requests can be modeled as a Pareto distribution. We also study the I/O load imbalance problems using I/O performance data collected from the Spider storage system.« less

Parallel processing for scientific computations

NASA Technical Reports Server (NTRS)

Alkhatib, Hasan S.

1995-01-01

The scope of this project dealt with the investigation of the requirements to support distributed computing of scientific computations over a cluster of cooperative workstations. Various experiments on computations for the solution of simultaneous linear equations were performed in the early phase of the project to gain experience in the general nature and requirements of scientific applications. A specification of a distributed integrated computing environment, DICE, based on a distributed shared memory communication paradigm has been developed and evaluated. The distributed shared memory model facilitates porting existing parallel algorithms that have been designed for shared memory multiprocessor systems to the new environment. The potential of this new environment is to provide supercomputing capability through the utilization of the aggregate power of workstations cooperating in a cluster interconnected via a local area network. Workstations, generally, do not have the computing power to tackle complex scientific applications, making them primarily useful for visualization, data reduction, and filtering as far as complex scientific applications are concerned. There is a tremendous amount of computing power that is left unused in a network of workstations. Very often a workstation is simply sitting idle on a desk. A set of tools can be developed to take advantage of this potential computing power to create a platform suitable for large scientific computations. The integration of several workstations into a logical cluster of distributed, cooperative, computing stations presents an alternative to shared memory multiprocessor systems. In this project we designed and evaluated such a system.

Fluid-structure interaction modeling of clusters of spacecraft parachutes with modified geometric porosity

NASA Astrophysics Data System (ADS)

Takizawa, Kenji; Tezduyar, Tayfun E.; Boben, Joseph; Kostov, Nikolay; Boswell, Cody; Buscher, Austin

2013-12-01

To increase aerodynamic performance, the geometric porosity of a ringsail spacecraft parachute canopy is sometimes increased, beyond the "rings" and "sails" with hundreds of "ring gaps" and "sail slits." This creates extra computational challenges for fluid-structure interaction (FSI) modeling of clusters of such parachutes, beyond those created by the lightness of the canopy structure, geometric complexities of hundreds of gaps and slits, and the contact between the parachutes of the cluster. In FSI computation of parachutes with such "modified geometric porosity," the flow through the "windows" created by the removal of the panels and the wider gaps created by the removal of the sails cannot be accurately modeled with the Homogenized Modeling of Geometric Porosity (HMGP), which was introduced to deal with the hundreds of gaps and slits. The flow needs to be actually resolved. All these computational challenges need to be addressed simultaneously in FSI modeling of clusters of spacecraft parachutes with modified geometric porosity. The core numerical technology is the Stabilized Space-Time FSI (SSTFSI) technique, and the contact between the parachutes is handled with the Surface-Edge-Node Contact Tracking (SENCT) technique. In the computations reported here, in addition to the SSTFSI and SENCT techniques and HMGP, we use the special techniques we have developed for removing the numerical spinning component of the parachute motion and for restoring the mesh integrity without a remesh. We present results for 2- and 3-parachute clusters with two different payload models.

Challenge Online Time Series Clustering For Demand Response A Theory to Break the ‘Curse of Dimensionality'

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

Pal, Ranjan; Chelmis, Charalampos; Aman, Saima

The advent of smart meters and advanced communication infrastructures catalyzes numerous smart grid applications such as dynamic demand response, and paves the way to solve challenging research problems in sustainable energy consumption. The space of solution possibilities are restricted primarily by the huge amount of generated data requiring considerable computational resources and efficient algorithms. To overcome this Big Data challenge, data clustering techniques have been proposed. Current approaches however do not scale in the face of the “increasing dimensionality” problem where a cluster point is represented by the entire customer consumption time series. To overcome this aspect we first rethinkmore » the way cluster points are created and designed, and then design an efficient online clustering technique for demand response (DR) in order to analyze high volume, high dimensional energy consumption time series data at scale, and on the fly. Our online algorithm is randomized in nature, and provides optimal performance guarantees in a computationally efficient manner. Unlike prior work we (i) study the consumption properties of the whole population simultaneously rather than developing individual models for each customer separately, claiming it to be a ‘killer’ approach that breaks the “curse of dimensionality” in online time series clustering, and (ii) provide tight performance guarantees in theory to validate our approach. Our insights are driven by the field of sociology, where collective behavior often emerges as the result of individual patterns and lifestyles.« less

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.

A web portal for hydrodynamical, cosmological simulations

NASA Astrophysics Data System (ADS)

Ragagnin, A.; Dolag, K.; Biffi, V.; Cadolle Bel, M.; Hammer, N. J.; Krukau, A.; Petkova, M.; Steinborn, D.

2017-07-01

This article describes a data centre hosting a web portal for accessing and sharing the output of large, cosmological, hydro-dynamical simulations with a broad scientific community. It also allows users to receive related scientific data products by directly processing the raw simulation data on a remote computing cluster. The data centre has a multi-layer structure: a web portal, a job control layer, a computing cluster and a HPC storage system. The outer layer enables users to choose an object from the simulations. Objects can be selected by visually inspecting 2D maps of the simulation data, by performing highly compounded and elaborated queries or graphically by plotting arbitrary combinations of properties. The user can run analysis tools on a chosen object. These services allow users to run analysis tools on the raw simulation data. The job control layer is responsible for handling and performing the analysis jobs, which are executed on a computing cluster. The innermost layer is formed by a HPC storage system which hosts the large, raw simulation data. The following services are available for the users: (I) CLUSTERINSPECT visualizes properties of member galaxies of a selected galaxy cluster; (II) SIMCUT returns the raw data of a sub-volume around a selected object from a simulation, containing all the original, hydro-dynamical quantities; (III) SMAC creates idealized 2D maps of various, physical quantities and observables of a selected object; (IV) PHOX generates virtual X-ray observations with specifications of various current and upcoming instruments.

Optimizing ion channel models using a parallel genetic algorithm on graphical processors.

PubMed

Ben-Shalom, Roy; Aviv, Amit; Razon, Benjamin; Korngreen, Alon

2012-01-01

We have recently shown that we can semi-automatically constrain models of voltage-gated ion channels by combining a stochastic search algorithm with ionic currents measured using multiple voltage-clamp protocols. Although numerically successful, this approach is highly demanding computationally, with optimization on a high performance Linux cluster typically lasting several days. To solve this computational bottleneck we converted our optimization algorithm for work on a graphical processing unit (GPU) using NVIDIA's CUDA. Parallelizing the process on a Fermi graphic computing engine from NVIDIA increased the speed ∼180 times over an application running on an 80 node Linux cluster, considerably reducing simulation times. This application allows users to optimize models for ion channel kinetics on a single, inexpensive, desktop "super computer," greatly reducing the time and cost of building models relevant to neuronal physiology. We also demonstrate that the point of algorithm parallelization is crucial to its performance. We substantially reduced computing time by solving the ODEs (Ordinary Differential Equations) so as to massively reduce memory transfers to and from the GPU. This approach may be applied to speed up other data intensive applications requiring iterative solutions of ODEs. Copyright © 2012 Elsevier B.V. All rights reserved.

Beating the tyranny of scale with a private cloud configured for Big Data

NASA Astrophysics Data System (ADS)

Lawrence, Bryan; Bennett, Victoria; Churchill, Jonathan; Juckes, Martin; Kershaw, Philip; Pepler, Sam; Pritchard, Matt; Stephens, Ag

2015-04-01

The Joint Analysis System, JASMIN, consists of a five significant hardware components: a batch computing cluster, a hypervisor cluster, bulk disk storage, high performance disk storage, and access to a tape robot. Each of the computing clusters consists of a heterogeneous set of servers, supporting a range of possible data analysis tasks - and a unique network environment makes it relatively trivial to migrate servers between the two clusters. The high performance disk storage will include the world's largest (publicly visible) deployment of the Panasas parallel disk system. Initially deployed in April 2012, JASMIN has already undergone two major upgrades, culminating in a system which by April 2015, will have in excess of 16 PB of disk and 4000 cores. Layered on the basic hardware are a range of services, ranging from managed services, such as the curated archives of the Centre for Environmental Data Archival or the data analysis environment for the National Centres for Atmospheric Science and Earth Observation, to a generic Infrastructure as a Service (IaaS) offering for the UK environmental science community. Here we present examples of some of the big data workloads being supported in this environment - ranging from data management tasks, such as checksumming 3 PB of data held in over one hundred million files, to science tasks, such as re-processing satellite observations with new algorithms, or calculating new diagnostics on petascale climate simulation outputs. We will demonstrate how the provision of a cloud environment closely coupled to a batch computing environment, all sharing the same high performance disk system allows massively parallel processing without the necessity to shuffle data excessively - even as it supports many different virtual communities, each with guaranteed performance. We will discuss the advantages of having a heterogeneous range of servers with available memory from tens of GB at the low end to (currently) two TB at the high end. There are some limitations of the JASMIN environment, the high performance disk environment is not fully available in the IaaS environment, and a planned ability to burst compute heavy jobs into the public cloud is not yet fully available. There are load balancing and performance issues that need to be understood. We will conclude with projections for future usage, and our plans to meet those requirements.

Workload Characterization of CFD Applications Using Partial Differential Equation Solvers

NASA Technical Reports Server (NTRS)

Waheed, Abdul; Yan, Jerry; Saini, Subhash (Technical Monitor)

1998-01-01

Workload characterization is used for modeling and evaluating of computing systems at different levels of detail. We present workload characterization for a class of Computational Fluid Dynamics (CFD) applications that solve Partial Differential Equations (PDEs). This workload characterization focuses on three high performance computing platforms: SGI Origin2000, EBM SP-2, a cluster of Intel Pentium Pro bases PCs. We execute extensive measurement-based experiments on these platforms to gather statistics of system resource usage, which results in workload characterization. Our workload characterization approach yields a coarse-grain resource utilization behavior that is being applied for performance modeling and evaluation of distributed high performance metacomputing systems. In addition, this study enhances our understanding of interactions between PDE solver workloads and high performance computing platforms and is useful for tuning these applications.

Efficient clustering aggregation based on data fragments.

PubMed

Wu, Ou; Hu, Weiming; Maybank, Stephen J; Zhu, Mingliang; Li, Bing

2012-06-01

Clustering aggregation, known as clustering ensembles, has emerged as a powerful technique for combining different clustering results to obtain a single better clustering. Existing clustering aggregation algorithms are applied directly to data points, in what is referred to as the point-based approach. The algorithms are inefficient if the number of data points is large. We define an efficient approach for clustering aggregation based on data fragments. In this fragment-based approach, a data fragment is any subset of the data that is not split by any of the clustering results. To establish the theoretical bases of the proposed approach, we prove that clustering aggregation can be performed directly on data fragments under two widely used goodness measures for clustering aggregation taken from the literature. Three new clustering aggregation algorithms are described. The experimental results obtained using several public data sets show that the new algorithms have lower computational complexity than three well-known existing point-based clustering aggregation algorithms (Agglomerative, Furthest, and LocalSearch); nevertheless, the new algorithms do not sacrifice the accuracy.

[Earth and Space Sciences Project Services for NASA HPCC

NASA Technical Reports Server (NTRS)

Merkey, Phillip

2002-01-01

This grant supported the effort to characterize the problem domain of the Earth Science Technology Office's Computational Technologies Project, to engage the Beowulf Cluster Computing Community as well as the High Performance Computing Research Community so that we can predict the applicability of said technologies to the scientific community represented by the CT project and formulate long term strategies to provide the computational resources necessary to attain the anticipated scientific objectives of the CT project. Specifically, the goal of the evaluation effort is to use the information gathered over the course of the Round-3 investigations to quantify the trends in scientific expectations, the algorithmic requirements and capabilities of high-performance computers to satisfy this anticipated need.

High-Performance Data Analysis Tools for Sun-Earth Connection Missions

NASA Technical Reports Server (NTRS)

Messmer, Peter

2011-01-01

The data analysis tool of choice for many Sun-Earth Connection missions is the Interactive Data Language (IDL) by ITT VIS. The increasing amount of data produced by these missions and the increasing complexity of image processing algorithms requires access to higher computing power. Parallel computing is a cost-effective way to increase the speed of computation, but algorithms oftentimes have to be modified to take advantage of parallel systems. Enhancing IDL to work on clusters gives scientists access to increased performance in a familiar programming environment. The goal of this project was to enable IDL applications to benefit from both computing clusters as well as graphics processing units (GPUs) for accelerating data analysis tasks. The tool suite developed in this project enables scientists now to solve demanding data analysis problems in IDL that previously required specialized software, and it allows them to be solved orders of magnitude faster than on conventional PCs. The tool suite consists of three components: (1) TaskDL, a software tool that simplifies the creation and management of task farms, collections of tasks that can be processed independently and require only small amounts of data communication; (2) mpiDL, a tool that allows IDL developers to use the Message Passing Interface (MPI) inside IDL for problems that require large amounts of data to be exchanged among multiple processors; and (3) GPULib, a tool that simplifies the use of GPUs as mathematical coprocessors from within IDL. mpiDL is unique in its support for the full MPI standard and its support of a broad range of MPI implementations. GPULib is unique in enabling users to take advantage of an inexpensive piece of hardware, possibly already installed in their computer, and achieve orders of magnitude faster execution time for numerically complex algorithms. TaskDL enables the simple setup and management of task farms on compute clusters. The products developed in this project have the potential to interact, so one can build a cluster of PCs, each equipped with a GPU, and use mpiDL to communicate between the nodes and GPULib to accelerate the computations on each node.

Cluster-state quantum computing enhanced by high-fidelity generalized measurements.

PubMed

Biggerstaff, D N; Kaltenbaek, R; Hamel, D R; Weihs, G; Rudolph, T; Resch, K J

2009-12-11

We introduce and implement a technique to extend the quantum computational power of cluster states by replacing some projective measurements with generalized quantum measurements (POVMs). As an experimental demonstration we fully realize an arbitrary three-qubit cluster computation by implementing a tunable linear-optical POVM, as well as fast active feedforward, on a two-qubit photonic cluster state. Over 206 different computations, the average output fidelity is 0.9832+/-0.0002; furthermore the error contribution from our POVM device and feedforward is only of O(10(-3)), less than some recent thresholds for fault-tolerant cluster computing.

Efficient and Flexible Climate Analysis with Python in a Cloud-Based Distributed Computing Framework

NASA Astrophysics Data System (ADS)

Gannon, C.

2017-12-01

As climate models become progressively more advanced, and spatial resolution further improved through various downscaling projects, climate projections at a local level are increasingly insightful and valuable. However, the raw size of climate datasets presents numerous hurdles for analysts wishing to develop customized climate risk metrics or perform site-specific statistical analysis. Four Twenty Seven, a climate risk consultancy, has implemented a Python-based distributed framework to analyze large climate datasets in the cloud. With the freedom afforded by efficiently processing these datasets, we are able to customize and continually develop new climate risk metrics using the most up-to-date data. Here we outline our process for using Python packages such as XArray and Dask to evaluate netCDF files in a distributed framework, StarCluster to operate in a cluster-computing environment, cloud computing services to access publicly hosted datasets, and how this setup is particularly valuable for generating climate change indicators and performing localized statistical analysis.

Computational evaluation of sub-nanometer cluster activity of singly exposed copper atom with various coordinative environment in catalytic CO2 transformation

NASA Astrophysics Data System (ADS)

Shanmugam, Ramasamy; Thamaraichelvan, Arunachalam; Ganesan, Tharumeya Kuppusamy; Viswanathan, Balasubramanian

2017-02-01

Metal cluster, at sub-nanometer level has a unique property in the activation of small molecules, in contrast to that of bulk surface. In the present work, singly exposed active site of copper metal cluster at sub-nanometer level was designed to arrive at the energy minimised configurations, binding energy, electrostatic potential map, frontier molecular orbitals and partial density of states. The ab initio molecular dynamics was carried out to probe the catalytic nature of the cluster. Further, the stability of the metal cluster and its catalytic activity in the electrochemical reduction of CO2 to CO were evaluated by means of computational hydrogen electrode via calculation of the free energy profile using DFT/B3LYP level of theory in vacuum. The activity of the cluster is ascertained from the fact that the copper atom, present in a two coordinative environment, performs a more selective conversion of CO2 to CO at an applied potential of -0.35 V which is comparatively lower than that of higher coordinative sites. The present study helps to design any sub-nano level metal catalyst for electrochemical reduction of CO2 to various value added chemicals.

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

Kim, Junghyun; Gangwon, Jo; Jaehoon, Jung

Applications written solely in OpenCL or CUDA cannot execute on a cluster as a whole. Most previous approaches that extend these programming models to clusters are based on a common idea: designating a centralized host node and coordinating the other nodes with the host for computation. However, the centralized host node is a serious performance bottleneck when the number of nodes is large. In this paper, we propose a scalable and distributed OpenCL framework called SnuCL-D for large-scale clusters. SnuCL-D's remote device virtualization provides an OpenCL application with an illusion that all compute devices in a cluster are confined inmore » a single node. To reduce the amount of control-message and data communication between nodes, SnuCL-D replicates the OpenCL host program execution and data in each node. We also propose a new OpenCL host API function and a queueing optimization technique that significantly reduce the overhead incurred by the previous centralized approaches. To show the effectiveness of SnuCL-D, we evaluate SnuCL-D with a microbenchmark and eleven benchmark applications on a large-scale CPU cluster and a medium-scale GPU cluster.« 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.

Sideloading - Ingestion of Large Point Clouds Into the Apache Spark Big Data Engine

NASA Astrophysics Data System (ADS)

Boehm, J.; Liu, K.; Alis, C.

2016-06-01

In the geospatial domain we have now reached the point where data volumes we handle have clearly grown beyond the capacity of most desktop computers. This is particularly true in the area of point cloud processing. It is therefore naturally lucrative to explore established big data frameworks for big geospatial data. The very first hurdle is the import of geospatial data into big data frameworks, commonly referred to as data ingestion. Geospatial data is typically encoded in specialised binary file formats, which are not naturally supported by the existing big data frameworks. Instead such file formats are supported by software libraries that are restricted to single CPU execution. We present an approach that allows the use of existing point cloud file format libraries on the Apache Spark big data framework. We demonstrate the ingestion of large volumes of point cloud data into a compute cluster. The approach uses a map function to distribute the data ingestion across the nodes of a cluster. We test the capabilities of the proposed method to load billions of points into a commodity hardware compute cluster and we discuss the implications on scalability and performance. The performance is benchmarked against an existing native Apache Spark data import implementation.

Evaluation of a micro-scale wind model's performance over realistic building clusters using wind tunnel experiments

NASA Astrophysics Data System (ADS)

Zhang, Ning; Du, Yunsong; Miao, Shiguang; Fang, Xiaoyi

2016-08-01

The simulation performance over complex building clusters of a wind simulation model (Wind Information Field Fast Analysis model, WIFFA) in a micro-scale air pollutant dispersion model system (Urban Microscale Air Pollution dispersion Simulation model, UMAPS) is evaluated using various wind tunnel experimental data including the CEDVAL (Compilation of Experimental Data for Validation of Micro-Scale Dispersion Models) wind tunnel experiment data and the NJU-FZ experiment data (Nanjing University-Fang Zhuang neighborhood wind tunnel experiment data). The results show that the wind model can reproduce the vortexes triggered by urban buildings well, and the flow patterns in urban street canyons and building clusters can also be represented. Due to the complex shapes of buildings and their distributions, the simulation deviations/discrepancies from the measurements are usually caused by the simplification of the building shapes and the determination of the key zone sizes. The computational efficiencies of different cases are also discussed in this paper. The model has a high computational efficiency compared to traditional numerical models that solve the Navier-Stokes equations, and can produce very high-resolution (1-5 m) wind fields of a complex neighborhood scale urban building canopy (~ 1 km ×1 km) in less than 3 min when run on a personal computer.

Hot Chips and Hot Interconnects for High End Computing Systems

NASA Technical Reports Server (NTRS)

Saini, Subhash

2005-01-01

I will discuss several processors: 1. The Cray proprietary processor used in the Cray X1; 2. The IBM Power 3 and Power 4 used in an IBM SP 3 and IBM SP 4 systems; 3. The Intel Itanium and Xeon, used in the SGI Altix systems and clusters respectively; 4. IBM System-on-a-Chip used in IBM BlueGene/L; 5. HP Alpha EV68 processor used in DOE ASCI Q cluster; 6. SPARC64 V processor, which is used in the Fujitsu PRIMEPOWER HPC2500; 7. An NEC proprietary processor, which is used in NEC SX-6/7; 8. Power 4+ processor, which is used in Hitachi SR11000; 9. NEC proprietary processor, which is used in Earth Simulator. The IBM POWER5 and Red Storm Computing Systems will also be discussed. The architectures of these processors will first be presented, followed by interconnection networks and a description of high-end computer systems based on these processors and networks. The performance of various hardware/programming model combinations will then be compared, based on latest NAS Parallel Benchmark results (MPI, OpenMP/HPF and hybrid (MPI + OpenMP). The tutorial will conclude with a discussion of general trends in the field of high performance computing, (quantum computing, DNA computing, cellular engineering, and neural networks).

Spatial cluster detection using dynamic programming.

PubMed

Sverchkov, Yuriy; Jiang, Xia; Cooper, Gregory F

2012-03-25

The task of spatial cluster detection involves finding spatial regions where some property deviates from the norm or the expected value. In a probabilistic setting this task can be expressed as finding a region where some event is significantly more likely than usual. Spatial cluster detection is of interest in fields such as biosurveillance, mining of astronomical data, military surveillance, and analysis of fMRI images. In almost all such applications we are interested both in the question of whether a cluster exists in the data, and if it exists, we are interested in finding the most accurate characterization of the cluster. We present a general dynamic programming algorithm for grid-based spatial cluster detection. The algorithm can be used for both Bayesian maximum a-posteriori (MAP) estimation of the most likely spatial distribution of clusters and Bayesian model averaging over a large space of spatial cluster distributions to compute the posterior probability of an unusual spatial clustering. The algorithm is explained and evaluated in the context of a biosurveillance application, specifically the detection and identification of Influenza outbreaks based on emergency department visits. A relatively simple underlying model is constructed for the purpose of evaluating the algorithm, and the algorithm is evaluated using the model and semi-synthetic test data. When compared to baseline methods, tests indicate that the new algorithm can improve MAP estimates under certain conditions: the greedy algorithm we compared our method to was found to be more sensitive to smaller outbreaks, while as the size of the outbreaks increases, in terms of area affected and proportion of individuals affected, our method overtakes the greedy algorithm in spatial precision and recall. The new algorithm performs on-par with baseline methods in the task of Bayesian model averaging. We conclude that the dynamic programming algorithm performs on-par with other available methods for spatial cluster detection and point to its low computational cost and extendability as advantages in favor of further research and use of the algorithm.

Spatial cluster detection using dynamic programming

PubMed Central

2012-01-01

Background The task of spatial cluster detection involves finding spatial regions where some property deviates from the norm or the expected value. In a probabilistic setting this task can be expressed as finding a region where some event is significantly more likely than usual. Spatial cluster detection is of interest in fields such as biosurveillance, mining of astronomical data, military surveillance, and analysis of fMRI images. In almost all such applications we are interested both in the question of whether a cluster exists in the data, and if it exists, we are interested in finding the most accurate characterization of the cluster. Methods We present a general dynamic programming algorithm for grid-based spatial cluster detection. The algorithm can be used for both Bayesian maximum a-posteriori (MAP) estimation of the most likely spatial distribution of clusters and Bayesian model averaging over a large space of spatial cluster distributions to compute the posterior probability of an unusual spatial clustering. The algorithm is explained and evaluated in the context of a biosurveillance application, specifically the detection and identification of Influenza outbreaks based on emergency department visits. A relatively simple underlying model is constructed for the purpose of evaluating the algorithm, and the algorithm is evaluated using the model and semi-synthetic test data. Results When compared to baseline methods, tests indicate that the new algorithm can improve MAP estimates under certain conditions: the greedy algorithm we compared our method to was found to be more sensitive to smaller outbreaks, while as the size of the outbreaks increases, in terms of area affected and proportion of individuals affected, our method overtakes the greedy algorithm in spatial precision and recall. The new algorithm performs on-par with baseline methods in the task of Bayesian model averaging. Conclusions We conclude that the dynamic programming algorithm performs on-par with other available methods for spatial cluster detection and point to its low computational cost and extendability as advantages in favor of further research and use of the algorithm. PMID:22443103

Parallel Calculations in LS-DYNA

NASA Astrophysics Data System (ADS)

Vartanovich Mkrtychev, Oleg; Aleksandrovich Reshetov, Andrey

2017-11-01

Nowadays, structural mechanics exhibits a trend towards numeric solutions being found for increasingly extensive and detailed tasks, which requires that capacities of computing systems be enhanced. Such enhancement can be achieved by different means. E.g., in case a computing system is represented by a workstation, its components can be replaced and/or extended (CPU, memory etc.). In essence, such modification eventually entails replacement of the entire workstation, i.e. replacement of certain components necessitates exchange of others (faster CPUs and memory devices require buses with higher throughput etc.). Special consideration must be given to the capabilities of modern video cards. They constitute powerful computing systems capable of running data processing in parallel. Interestingly, the tools originally designed to render high-performance graphics can be applied for solving problems not immediately related to graphics (CUDA, OpenCL, Shaders etc.). However, not all software suites utilize video cards’ capacities. Another way to increase capacity of a computing system is to implement a cluster architecture: to add cluster nodes (workstations) and to increase the network communication speed between the nodes. The advantage of this approach is extensive growth due to which a quite powerful system can be obtained by combining not particularly powerful nodes. Moreover, separate nodes may possess different capacities. This paper considers the use of a clustered computing system for solving problems of structural mechanics with LS-DYNA software. To establish a range of dependencies a mere 2-node cluster has proven sufficient.

Online clustering algorithms for radar emitter classification.

PubMed

Liu, Jun; Lee, Jim P Y; Senior; Li, Lingjie; Luo, Zhi-Quan; Wong, K Max

2005-08-01

Radar emitter classification is a special application of data clustering for classifying unknown radar emitters from received radar pulse samples. The main challenges of this task are the high dimensionality of radar pulse samples, small sample group size, and closely located radar pulse clusters. In this paper, two new online clustering algorithms are developed for radar emitter classification: One is model-based using the Minimum Description Length (MDL) criterion and the other is based on competitive learning. Computational complexity is analyzed for each algorithm and then compared. Simulation results show the superior performance of the model-based algorithm over competitive learning in terms of better classification accuracy, flexibility, and stability.

Quantum-Size Dependence of the Energy for Vacancy Formation in Charged Small Metal Clusters. Drop Model

NASA Astrophysics Data System (ADS)

Pogosov, V. V.; Reva, V. I.

2018-04-01

Self-consistent computations of the monovacancy formation energy are performed for Na N , Mg N , and Al N (12 < N ≤ 168) spherical clusters in the drop model for stable jelly. Scenarios of the Schottky vacancy formation and "bubble vacancy blowing" are considered. It is shown that the asymptotic behavior of the size dependences of the energy for the vacancy formation by these two mechanisms is different and the difference between the characteristics of a charged and neutral cluster is entirely determined by the difference between the ionization potentials of clusters and the energies of electron attachment to them.

TethysCluster: A comprehensive approach for harnessing cloud resources for hydrologic modeling

NASA Astrophysics Data System (ADS)

Nelson, J.; Jones, N.; Ames, D. P.

2015-12-01

Advances in water resources modeling are improving the information that can be supplied to support decisions affecting the safety and sustainability of society. However, as water resources models become more sophisticated and data-intensive they require more computational power to run. Purchasing and maintaining the computing facilities needed to support certain modeling tasks has been cost-prohibitive for many organizations. With the advent of the cloud, the computing resources needed to address this challenge are now available and cost-effective, yet there still remains a significant technical barrier to leverage these resources. This barrier inhibits many decision makers and even trained engineers from taking advantage of the best science and tools available. Here we present the Python tools TethysCluster and CondorPy, that have been developed to lower the barrier to model computation in the cloud by providing (1) programmatic access to dynamically scalable computing resources, (2) a batch scheduling system to queue and dispatch the jobs to the computing resources, (3) data management for job inputs and outputs, and (4) the ability to dynamically create, submit, and monitor computing jobs. These Python tools leverage the open source, computing-resource management, and job management software, HTCondor, to offer a flexible and scalable distributed-computing environment. While TethysCluster and CondorPy can be used independently to provision computing resources and perform large modeling tasks, they have also been integrated into Tethys Platform, a development platform for water resources web apps, to enable computing support for modeling workflows and decision-support systems deployed as web apps.

Near-Edge X-ray Absorption Fine Structure within Multilevel Coupled Cluster Theory.

PubMed

Myhre, Rolf H; Coriani, Sonia; Koch, Henrik

2016-06-14

Core excited states are challenging to calculate, mainly because they are embedded in a manifold of high-energy valence-excited states. However, their locality makes their determination ideal for local correlation methods. In this paper, we demonstrate the performance of multilevel coupled cluster theory in computing core spectra both within the core-valence separated and the asymmetric Lanczos implementations of coupled cluster linear response theory. We also propose a visualization tool to analyze the excitations using the difference between the ground-state and excited-state electron densities.

Region Templates: Data Representation and Management for High-Throughput Image Analysis

PubMed Central

Pan, Tony; Kurc, Tahsin; Kong, Jun; Cooper, Lee; Klasky, Scott; Saltz, Joel

2015-01-01

We introduce a region template abstraction and framework for the efficient storage, management and processing of common data types in analysis of large datasets of high resolution images on clusters of hybrid computing nodes. The region template abstraction provides a generic container template for common data structures, such as points, arrays, regions, and object sets, within a spatial and temporal bounding box. It allows for different data management strategies and I/O implementations, while providing a homogeneous, unified interface to applications for data storage and retrieval. A region template application is represented as a hierarchical dataflow in which each computing stage may be represented as another dataflow of finer-grain tasks. The execution of the application is coordinated by a runtime system that implements optimizations for hybrid machines, including performance-aware scheduling for maximizing the utilization of computing devices and techniques to reduce the impact of data transfers between CPUs and GPUs. An experimental evaluation on a state-of-the-art hybrid cluster using a microscopy imaging application shows that the abstraction adds negligible overhead (about 3%) and achieves good scalability and high data transfer rates. Optimizations in a high speed disk based storage implementation of the abstraction to support asynchronous data transfers and computation result in an application performance gain of about 1.13×. Finally, a processing rate of 11,730 4K×4K tiles per minute was achieved for the microscopy imaging application on a cluster with 100 nodes (300 GPUs and 1,200 CPU cores). This computation rate enables studies with very large datasets. PMID:26139953

UQlust: combining profile hashing with linear-time ranking for efficient clustering and analysis of big macromolecular data.

PubMed

Adamczak, Rafal; Meller, Jarek

2016-12-28

Advances in computing have enabled current protein and RNA structure prediction and molecular simulation methods to dramatically increase their sampling of conformational spaces. The quickly growing number of experimentally resolved structures, and databases such as the Protein Data Bank, also implies large scale structural similarity analyses to retrieve and classify macromolecular data. Consequently, the computational cost of structure comparison and clustering for large sets of macromolecular structures has become a bottleneck that necessitates further algorithmic improvements and development of efficient software solutions. uQlust is a versatile and easy-to-use tool for ultrafast ranking and clustering of macromolecular structures. uQlust makes use of structural profiles of proteins and nucleic acids, while combining a linear-time algorithm for implicit comparison of all pairs of models with profile hashing to enable efficient clustering of large data sets with a low memory footprint. In addition to ranking and clustering of large sets of models of the same protein or RNA molecule, uQlust can also be used in conjunction with fragment-based profiles in order to cluster structures of arbitrary length. For example, hierarchical clustering of the entire PDB using profile hashing can be performed on a typical laptop, thus opening an avenue for structural explorations previously limited to dedicated resources. The uQlust package is freely available under the GNU General Public License at https://github.com/uQlust . uQlust represents a drastic reduction in the computational complexity and memory requirements with respect to existing clustering and model quality assessment methods for macromolecular structure analysis, while yielding results on par with traditional approaches for both proteins and RNAs.

Creating a Parallel Version of VisIt for Microsoft Windows

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

Whitlock, B J; Biagas, K S; Rawson, P L

2011-12-07

VisIt is a popular, free interactive parallel visualization and analysis tool for scientific data. Users can quickly generate visualizations from their data, animate them through time, manipulate them, and save the resulting images or movies for presentations. VisIt was designed from the ground up to work on many scales of computers from modest desktops up to massively parallel clusters. VisIt is comprised of a set of cooperating programs. All programs can be run locally or in client/server mode in which some run locally and some run remotely on compute clusters. The VisIt program most able to harness today's computing powermore » is the VisIt compute engine. The compute engine is responsible for reading simulation data from disk, processing it, and sending results or images back to the VisIt viewer program. In a parallel environment, the compute engine runs several processes, coordinating using the Message Passing Interface (MPI) library. Each MPI process reads some subset of the scientific data and filters the data in various ways to create useful visualizations. By using MPI, VisIt has been able to scale well into the thousands of processors on large computers such as dawn and graph at LLNL. The advent of multicore CPU's has made parallelism the 'new' way to achieve increasing performance. With today's computers having at least 2 cores and in many cases up to 8 and beyond, it is more important than ever to deploy parallel software that can use that computing power not only on clusters but also on the desktop. We have created a parallel version of VisIt for Windows that uses Microsoft's MPI implementation (MSMPI) to process data in parallel on the Windows desktop as well as on a Windows HPC cluster running Microsoft Windows Server 2008. Initial desktop parallel support for Windows was deployed in VisIt 2.4.0. Windows HPC cluster support has been completed and will appear in the VisIt 2.5.0 release. We plan to continue supporting parallel VisIt on Windows so our users will be able to take full advantage of their multicore resources.« less

A clustering package for nucleotide sequences using Laplacian Eigenmaps and Gaussian Mixture Model.

PubMed

Bruneau, Marine; Mottet, Thierry; Moulin, Serge; Kerbiriou, Maël; Chouly, Franz; Chretien, Stéphane; Guyeux, Christophe

2018-02-01

In this article, a new Python package for nucleotide sequences clustering is proposed. This package, freely available on-line, implements a Laplacian eigenmap embedding and a Gaussian Mixture Model for DNA clustering. It takes nucleotide sequences as input, and produces the optimal number of clusters along with a relevant visualization. Despite the fact that we did not optimise the computational speed, our method still performs reasonably well in practice. Our focus was mainly on data analytics and accuracy and as a result, our approach outperforms the state of the art, even in the case of divergent sequences. Furthermore, an a priori knowledge on the number of clusters is not required here. For the sake of illustration, this method is applied on a set of 100 DNA sequences taken from the mitochondrially encoded NADH dehydrogenase 3 (ND3) gene, extracted from a collection of Platyhelminthes and Nematoda species. The resulting clusters are tightly consistent with the phylogenetic tree computed using a maximum likelihood approach on gene alignment. They are coherent too with the NCBI taxonomy. Further test results based on synthesized data are then provided, showing that the proposed approach is better able to recover the clusters than the most widely used software, namely Cd-hit-est and BLASTClust. Copyright © 2017 Elsevier Ltd. All rights reserved.

A Spatial Division Clustering Method and Low Dimensional Feature Extraction Technique Based Indoor Positioning System

PubMed Central

Mo, Yun; Zhang, Zhongzhao; Meng, Weixiao; Ma, Lin; Wang, Yao

2014-01-01

Indoor positioning systems based on the fingerprint method are widely used due to the large number of existing devices with a wide range of coverage. However, extensive positioning regions with a massive fingerprint database may cause high computational complexity and error margins, therefore clustering methods are widely applied as a solution. However, traditional clustering methods in positioning systems can only measure the similarity of the Received Signal Strength without being concerned with the continuity of physical coordinates. Besides, outage of access points could result in asymmetric matching problems which severely affect the fine positioning procedure. To solve these issues, in this paper we propose a positioning system based on the Spatial Division Clustering (SDC) method for clustering the fingerprint dataset subject to physical distance constraints. With the Genetic Algorithm and Support Vector Machine techniques, SDC can achieve higher coarse positioning accuracy than traditional clustering algorithms. In terms of fine localization, based on the Kernel Principal Component Analysis method, the proposed positioning system outperforms its counterparts based on other feature extraction methods in low dimensionality. Apart from balancing online matching computational burden, the new positioning system exhibits advantageous performance on radio map clustering, and also shows better robustness and adaptability in the asymmetric matching problem aspect. PMID:24451470

Institutional Computing Executive Group Review of Multi-programmatic & Institutional Computing, Fiscal Year 2005 and 2006

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

Langer, S; Rotman, D; Schwegler, E

The Institutional Computing Executive Group (ICEG) review of FY05-06 Multiprogrammatic and Institutional Computing (M and IC) activities is presented in the attached report. In summary, we find that the M and IC staff does an outstanding job of acquiring and supporting a wide range of institutional computing resources to meet the programmatic and scientific goals of LLNL. The responsiveness and high quality of support given to users and the programs investing in M and IC reflects the dedication and skill of the M and IC staff. M and IC has successfully managed serial capacity, parallel capacity, and capability computing resources.more » Serial capacity computing supports a wide range of scientific projects which require access to a few high performance processors within a shared memory computer. Parallel capacity computing supports scientific projects that require a moderate number of processors (up to roughly 1000) on a parallel computer. Capability computing supports parallel jobs that push the limits of simulation science. M and IC has worked closely with Stockpile Stewardship, and together they have made LLNL a premier institution for computational and simulation science. Such a standing is vital to the continued success of laboratory science programs and to the recruitment and retention of top scientists. This report provides recommendations to build on M and IC's accomplishments and improve simulation capabilities at LLNL. We recommend that institution fully fund (1) operation of the atlas cluster purchased in FY06 to support a few large projects; (2) operation of the thunder and zeus clusters to enable 'mid-range' parallel capacity simulations during normal operation and a limited number of large simulations during dedicated application time; (3) operation of the new yana cluster to support a wide range of serial capacity simulations; (4) improvements to the reliability and performance of the Lustre parallel file system; (5) support for the new GDO petabyte-class storage facility on the green network for use in data intensive external collaborations; and (6) continued support for visualization and other methods for analyzing large simulations. We also recommend that M and IC begin planning in FY07 for the next upgrade of its parallel clusters. LLNL investments in M and IC have resulted in a world-class simulation capability leading to innovative science. We thank the LLNL management for its continued support and thank the M and IC staff for its vision and dedicated efforts to make it all happen.« less

Size-guided multi-seed heuristic method for geometry optimization of clusters: Application to benzene clusters.

PubMed

Takeuchi, Hiroshi

2018-05-08

Since searching for the global minimum on the potential energy surface of a cluster is very difficult, many geometry optimization methods have been proposed, in which initial geometries are randomly generated and subsequently improved with different algorithms. In this study, a size-guided multi-seed heuristic method is developed and applied to benzene clusters. It produces initial configurations of the cluster with n molecules from the lowest-energy configurations of the cluster with n - 1 molecules (seeds). The initial geometries are further optimized with the geometrical perturbations previously used for molecular clusters. These steps are repeated until the size n satisfies a predefined one. The method locates putative global minima of benzene clusters with up to 65 molecules. The performance of the method is discussed using the computational cost, rates to locate the global minima, and energies of initial geometries. © 2018 Wiley Periodicals, Inc. © 2018 Wiley Periodicals, Inc.

Merging K-means with hierarchical clustering for identifying general-shaped groups.

PubMed

Peterson, Anna D; Ghosh, Arka P; Maitra, Ranjan

2018-01-01

Clustering partitions a dataset such that observations placed together in a group are similar but different from those in other groups. Hierarchical and K -means clustering are two approaches but have different strengths and weaknesses. For instance, hierarchical clustering identifies groups in a tree-like structure but suffers from computational complexity in large datasets while K -means clustering is efficient but designed to identify homogeneous spherically-shaped clusters. We present a hybrid non-parametric clustering approach that amalgamates the two methods to identify general-shaped clusters and that can be applied to larger datasets. Specifically, we first partition the dataset into spherical groups using K -means. We next merge these groups using hierarchical methods with a data-driven distance measure as a stopping criterion. Our proposal has the potential to reveal groups with general shapes and structure in a dataset. We demonstrate good performance on several simulated and real datasets.

Perspective: Size selected clusters for catalysis and electrochemistry

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

Halder, Avik; Curtiss, Larry A.; Fortunelli, Alessandro

We report that size-selected clusters containing a handful of atoms may possess noble catalytic properties different from nano-sized or bulk catalysts. Size- and composition-selected clusters can also serve as models of the catalytic active site, where an addition or removal of a single atom can have a dramatic effect on their activity and selectivity. In this Perspective, we provide an overview of studies performed under both ultra-high vacuum and realistic reaction conditions aimed at the interrogation, characterization and understanding of the performance of supported size-selected clusters in heterogeneous and electrochemical reactions, which address the effects of cluster size, cluster composition,more » cluster-support interactions and reaction conditions, the key parameters for the understanding and control of catalyst functionality. Computational modelling based on density functional theory sampling of local minima and energy barriers or ab initio Molecular Dynamics simulations is an integral part of this research by providing fundamental understanding of the catalytic processes at the atomic level, as well as by predicting new materials compositions which can be validated in experiments. Lastly, we discuss approaches which aim at the scale up of the production of well-defined clusters for use in real world applications.« less

Perspective: Size selected clusters for catalysis and electrochemistry

DOE PAGES

Halder, Avik; Curtiss, Larry A.; Fortunelli, Alessandro; ...

2018-03-15

We report that size-selected clusters containing a handful of atoms may possess noble catalytic properties different from nano-sized or bulk catalysts. Size- and composition-selected clusters can also serve as models of the catalytic active site, where an addition or removal of a single atom can have a dramatic effect on their activity and selectivity. In this Perspective, we provide an overview of studies performed under both ultra-high vacuum and realistic reaction conditions aimed at the interrogation, characterization and understanding of the performance of supported size-selected clusters in heterogeneous and electrochemical reactions, which address the effects of cluster size, cluster composition,more » cluster-support interactions and reaction conditions, the key parameters for the understanding and control of catalyst functionality. Computational modelling based on density functional theory sampling of local minima and energy barriers or ab initio Molecular Dynamics simulations is an integral part of this research by providing fundamental understanding of the catalytic processes at the atomic level, as well as by predicting new materials compositions which can be validated in experiments. Lastly, we discuss approaches which aim at the scale up of the production of well-defined clusters for use in real world applications.« less

Perspective: Size selected clusters for catalysis and electrochemistry

NASA Astrophysics Data System (ADS)

Halder, Avik; Curtiss, Larry A.; Fortunelli, Alessandro; Vajda, Stefan

2018-03-01

Size-selected clusters containing a handful of atoms may possess noble catalytic properties different from nano-sized or bulk catalysts. Size- and composition-selected clusters can also serve as models of the catalytic active site, where an addition or removal of a single atom can have a dramatic effect on their activity and selectivity. In this perspective, we provide an overview of studies performed under both ultra-high vacuum and realistic reaction conditions aimed at the interrogation, characterization, and understanding of the performance of supported size-selected clusters in heterogeneous and electrochemical reactions, which address the effects of cluster size, cluster composition, cluster-support interactions, and reaction conditions, the key parameters for the understanding and control of catalyst functionality. Computational modeling based on density functional theory sampling of local minima and energy barriers or ab initio molecular dynamics simulations is an integral part of this research by providing fundamental understanding of the catalytic processes at the atomic level, as well as by predicting new materials compositions which can be validated in experiments. Finally, we discuss approaches which aim at the scale up of the production of well-defined clusters for use in real world applications.

A computational system for lattice QCD with overlap Dirac quarks

NASA Astrophysics Data System (ADS)

Chiu, Ting-Wai; Hsieh, Tung-Han; Huang, Chao-Hsi; Huang, Tsung-Ren

2003-05-01

We outline the essential features of a Linux PC cluster which is now being developed at National Taiwan University, and discuss how to optimize its hardware and software for lattice QCD with overlap Dirac quarks. At present, the cluster constitutes of 30 nodes, with each node consisting of one Pentium 4 processor (1.6/2.0 GHz), one Gbyte of PC800 RDRAM, one 40/80 Gbyte hard disk, and a network card. The speed of this system is estimated to be 30 Gflops, and its price/performance ratio is better than $1.0/Mflops for 64-bit (double precision) computations in quenched lattice QCD with overlap Dirac quarks.

Membership determination of open clusters based on a spectral clustering method

NASA Astrophysics Data System (ADS)

Gao, Xin-Hua

2018-06-01

We present a spectral clustering (SC) method aimed at segregating reliable members of open clusters in multi-dimensional space. The SC method is a non-parametric clustering technique that performs cluster division using eigenvectors of the similarity matrix; no prior knowledge of the clusters is required. This method is more flexible in dealing with multi-dimensional data compared to other methods of membership determination. We use this method to segregate the cluster members of five open clusters (Hyades, Coma Ber, Pleiades, Praesepe, and NGC 188) in five-dimensional space; fairly clean cluster members are obtained. We find that the SC method can capture a small number of cluster members (weak signal) from a large number of field stars (heavy noise). Based on these cluster members, we compute the mean proper motions and distances for the Hyades, Coma Ber, Pleiades, and Praesepe clusters, and our results are in general quite consistent with the results derived by other authors. The test results indicate that the SC method is highly suitable for segregating cluster members of open clusters based on high-precision multi-dimensional astrometric data such as Gaia data.

Job Management Requirements for NAS Parallel Systems and Clusters

NASA Technical Reports Server (NTRS)

Saphir, William; Tanner, Leigh Ann; Traversat, Bernard

1995-01-01

A job management system is a critical component of a production supercomputing environment, permitting oversubscribed resources to be shared fairly and efficiently. Job management systems that were originally designed for traditional vector supercomputers are not appropriate for the distributed-memory parallel supercomputers that are becoming increasingly important in the high performance computing industry. Newer job management systems offer new functionality but do not solve fundamental problems. We address some of the main issues in resource allocation and job scheduling we have encountered on two parallel computers - a 160-node IBM SP2 and a cluster of 20 high performance workstations located at the Numerical Aerodynamic Simulation facility. We describe the requirements for resource allocation and job management that are necessary to provide a production supercomputing environment on these machines, prioritizing according to difficulty and importance, and advocating a return to fundamental issues.

Autogrid-based clustering of kinases: selection of representative conformations for docking purposes.

PubMed

Marzaro, Giovanni; Ferrarese, Alessandro; Chilin, Adriana

2014-08-01

The selection of the most appropriate protein conformation is a crucial aspect in molecular docking experiments. In order to reduce the errors arising from the use of a single protein conformation, several authors suggest the use of several tridimensional structures for the target. However, the selection of the most appropriate protein conformations still remains a challenging goal. The protein 3D-structures selection is mainly performed based on pairwise root-mean-square-deviation (RMSD) values computation, followed by hierarchical clustering. Herein we report an alternative strategy, based on the computation of only two atom affinity map for each protein conformation, followed by multivariate analysis and hierarchical clustering. This methodology was applied on seven different kinases of pharmaceutical interest. The comparison with the classical RMSD-based strategy was based on cross-docking of co-crystallized ligands. In the case of epidermal growth factor receptor kinase, also the docking performance on 220 known ligands were evaluated, followed by 3D-QSAR studies. In all the cases, the herein proposed methodology outperformed the RMSD-based one.

Potential Application of a Graphical Processing Unit to Parallel Computations in the NUBEAM Code

NASA Astrophysics Data System (ADS)

Payne, J.; McCune, D.; Prater, R.

2010-11-01

NUBEAM is a comprehensive computational Monte Carlo based model for neutral beam injection (NBI) in tokamaks. NUBEAM computes NBI-relevant profiles in tokamak plasmas by tracking the deposition and the slowing of fast ions. At the core of NUBEAM are vector calculations used to track fast ions. These calculations have recently been parallelized to run on MPI clusters. However, cost and interlink bandwidth limit the ability to fully parallelize NUBEAM on an MPI cluster. Recent implementation of double precision capabilities for Graphical Processing Units (GPUs) presents a cost effective and high performance alternative or complement to MPI computation. Commercially available graphics cards can achieve up to 672 GFLOPS double precision and can handle hundreds of thousands of threads. The ability to execute at least one thread per particle simultaneously could significantly reduce the execution time and the statistical noise of NUBEAM. Progress on implementation on a GPU will be presented.

Computer aided detection of clusters of microcalcifications on full field digital mammograms

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

Ge Jun; Sahiner, Berkman; Hadjiiski, Lubomir M.

2006-08-15

We are developing a computer-aided detection (CAD) system to identify microcalcification clusters (MCCs) automatically on full field digital mammograms (FFDMs). The CAD system includes six stages: preprocessing; image enhancement; segmentation of microcalcification candidates; false positive (FP) reduction for individual microcalcifications; regional clustering; and FP reduction for clustered microcalcifications. At the stage of FP reduction for individual microcalcifications, a truncated sum-of-squares error function was used to improve the efficiency and robustness of the training of an artificial neural network in our CAD system for FFDMs. At the stage of FP reduction for clustered microcalcifications, morphological features and features derived from themore » artificial neural network outputs were extracted from each cluster. Stepwise linear discriminant analysis (LDA) was used to select the features. An LDA classifier was then used to differentiate clustered microcalcifications from FPs. A data set of 96 cases with 192 images was collected at the University of Michigan. This data set contained 96 MCCs, of which 28 clusters were proven by biopsy to be malignant and 68 were proven to be benign. The data set was separated into two independent data sets for training and testing of the CAD system in a cross-validation scheme. When one data set was used to train and validate the convolution neural network (CNN) in our CAD system, the other data set was used to evaluate the detection performance. With the use of a truncated error metric, the training of CNN could be accelerated and the classification performance was improved. The CNN in combination with an LDA classifier could substantially reduce FPs with a small tradeoff in sensitivity. By using the free-response receiver operating characteristic methodology, it was found that our CAD system can achieve a cluster-based sensitivity of 70, 80, and 90 % at 0.21, 0.61, and 1.49 FPs/image, respectively. For case-based performance evaluation, a sensitivity of 70, 80, and 90 % can be achieved at 0.07, 0.17, and 0.65 FPs/image, respectively. We also used a data set of 216 mammograms negative for clustered microcalcifications to further estimate the FP rate of our CAD system. The corresponding FP rates were 0.15, 0.31, and 0.86 FPs/image for cluster-based detection when negative mammograms were used for estimation of FP rates.« less

Demonstration of measurement-only blind quantum computing

NASA Astrophysics Data System (ADS)

Greganti, Chiara; Roehsner, Marie-Christine; Barz, Stefanie; Morimae, Tomoyuki; Walther, Philip

2016-01-01

Blind quantum computing allows for secure cloud networks of quasi-classical clients and a fully fledged quantum server. Recently, a new protocol has been proposed, which requires a client to perform only measurements. We demonstrate a proof-of-principle implementation of this measurement-only blind quantum computing, exploiting a photonic setup to generate four-qubit cluster states for computation and verification. Feasible technological requirements for the client and the device-independent blindness make this scheme very applicable for future secure quantum networks.

Heads in the Cloud: A Primer on Neuroimaging Applications of High Performance Computing.

PubMed

Shatil, Anwar S; Younas, Sohail; Pourreza, Hossein; Figley, Chase R

2015-01-01

With larger data sets and more sophisticated analyses, it is becoming increasingly common for neuroimaging researchers to push (or exceed) the limitations of standalone computer workstations. Nonetheless, although high-performance computing platforms such as clusters, grids and clouds are already in routine use by a small handful of neuroimaging researchers to increase their storage and/or computational power, the adoption of such resources by the broader neuroimaging community remains relatively uncommon. Therefore, the goal of the current manuscript is to: 1) inform prospective users about the similarities and differences between computing clusters, grids and clouds; 2) highlight their main advantages; 3) discuss when it may (and may not) be advisable to use them; 4) review some of their potential problems and barriers to access; and finally 5) give a few practical suggestions for how interested new users can start analyzing their neuroimaging data using cloud resources. Although the aim of cloud computing is to hide most of the complexity of the infrastructure management from end-users, we recognize that this can still be an intimidating area for cognitive neuroscientists, psychologists, neurologists, radiologists, and other neuroimaging researchers lacking a strong computational background. Therefore, with this in mind, we have aimed to provide a basic introduction to cloud computing in general (including some of the basic terminology, computer architectures, infrastructure and service models, etc.), a practical overview of the benefits and drawbacks, and a specific focus on how cloud resources can be used for various neuroimaging applications.

Heads in the Cloud: A Primer on Neuroimaging Applications of High Performance Computing

PubMed Central

Shatil, Anwar S.; Younas, Sohail; Pourreza, Hossein; Figley, Chase R.

2015-01-01

With larger data sets and more sophisticated analyses, it is becoming increasingly common for neuroimaging researchers to push (or exceed) the limitations of standalone computer workstations. Nonetheless, although high-performance computing platforms such as clusters, grids and clouds are already in routine use by a small handful of neuroimaging researchers to increase their storage and/or computational power, the adoption of such resources by the broader neuroimaging community remains relatively uncommon. Therefore, the goal of the current manuscript is to: 1) inform prospective users about the similarities and differences between computing clusters, grids and clouds; 2) highlight their main advantages; 3) discuss when it may (and may not) be advisable to use them; 4) review some of their potential problems and barriers to access; and finally 5) give a few practical suggestions for how interested new users can start analyzing their neuroimaging data using cloud resources. Although the aim of cloud computing is to hide most of the complexity of the infrastructure management from end-users, we recognize that this can still be an intimidating area for cognitive neuroscientists, psychologists, neurologists, radiologists, and other neuroimaging researchers lacking a strong computational background. Therefore, with this in mind, we have aimed to provide a basic introduction to cloud computing in general (including some of the basic terminology, computer architectures, infrastructure and service models, etc.), a practical overview of the benefits and drawbacks, and a specific focus on how cloud resources can be used for various neuroimaging applications. PMID:27279746

Developing the fuzzy c-means clustering algorithm based on maximum entropy for multitarget tracking in a cluttered environment

NASA Astrophysics Data System (ADS)

Chen, Xiao; Li, Yaan; Yu, Jing; Li, Yuxing

2018-01-01

For fast and more effective implementation of tracking multiple targets in a cluttered environment, we propose a multiple targets tracking (MTT) algorithm called maximum entropy fuzzy c-means clustering joint probabilistic data association that combines fuzzy c-means clustering and the joint probabilistic data association (PDA) algorithm. The algorithm uses the membership value to express the probability of the target originating from measurement. The membership value is obtained through fuzzy c-means clustering objective function optimized by the maximum entropy principle. When considering the effect of the public measurement, we use a correction factor to adjust the association probability matrix to estimate the state of the target. As this algorithm avoids confirmation matrix splitting, it can solve the high computational load problem of the joint PDA algorithm. The results of simulations and analysis conducted for tracking neighbor parallel targets and cross targets in a different density cluttered environment show that the proposed algorithm can realize MTT quickly and efficiently in a cluttered environment. Further, the performance of the proposed algorithm remains constant with increasing process noise variance. The proposed algorithm has the advantages of efficiency and low computational load, which can ensure optimum performance when tracking multiple targets in a dense cluttered environment.

Parameters that affect parallel processing for computational electromagnetic simulation codes on high performance computing clusters

NASA Astrophysics Data System (ADS)

Moon, Hongsik

What is the impact of multicore and associated advanced technologies on computational software for science? Most researchers and students have multicore laptops or desktops for their research and they need computing power to run computational software packages. Computing power was initially derived from Central Processing Unit (CPU) clock speed. That changed when increases in clock speed became constrained by power requirements. Chip manufacturers turned to multicore CPU architectures and associated technological advancements to create the CPUs for the future. Most software applications benefited by the increased computing power the same way that increases in clock speed helped applications run faster. However, for Computational ElectroMagnetics (CEM) software developers, this change was not an obvious benefit - it appeared to be a detriment. Developers were challenged to find a way to correctly utilize the advancements in hardware so that their codes could benefit. The solution was parallelization and this dissertation details the investigation to address these challenges. Prior to multicore CPUs, advanced computer technologies were compared with the performance using benchmark software and the metric was FLoting-point Operations Per Seconds (FLOPS) which indicates system performance for scientific applications that make heavy use of floating-point calculations. Is FLOPS an effective metric for parallelized CEM simulation tools on new multicore system? Parallel CEM software needs to be benchmarked not only by FLOPS but also by the performance of other parameters related to type and utilization of the hardware, such as CPU, Random Access Memory (RAM), hard disk, network, etc. The codes need to be optimized for more than just FLOPs and new parameters must be included in benchmarking. In this dissertation, the parallel CEM software named High Order Basis Based Integral Equation Solver (HOBBIES) is introduced. This code was developed to address the needs of the changing computer hardware platforms in order to provide fast, accurate and efficient solutions to large, complex electromagnetic problems. The research in this dissertation proves that the performance of parallel code is intimately related to the configuration of the computer hardware and can be maximized for different hardware platforms. To benchmark and optimize the performance of parallel CEM software, a variety of large, complex projects are created and executed on a variety of computer platforms. The computer platforms used in this research are detailed in this dissertation. The projects run as benchmarks are also described in detail and results are presented. The parameters that affect parallel CEM software on High Performance Computing Clusters (HPCC) are investigated. This research demonstrates methods to maximize the performance of parallel CEM software code.

Efficient generation of low-energy folded states of a model protein

NASA Astrophysics Data System (ADS)

Gordon, Heather L.; Kwan, Wai Kei; Gong, Chunhang; Larrass, Stefan; Rothstein, Stuart M.

2003-01-01

A number of short simulated annealing runs are performed on a highly-frustrated 46-"residue" off-lattice model protein. We perform, in an iterative fashion, a principal component analysis of the 946 nonbonded interbead distances, followed by two varieties of cluster analyses: hierarchical and k-means clustering. We identify several distinct sets of conformations with reasonably consistent cluster membership. Nonbonded distance constraints are derived for each cluster and are employed within a distance geometry approach to generate many new conformations, previously unidentified by the simulated annealing experiments. Subsequent analyses suggest that these new conformations are members of the parent clusters from which they were generated. Furthermore, several novel, previously unobserved structures with low energy were uncovered, augmenting the ensemble of simulated annealing results, and providing a complete distribution of low-energy states. The computational cost of this approach to generating low-energy conformations is small when compared to the expense of further Monte Carlo simulated annealing runs.

Fast Multipole Methods for Three-Dimensional N-body Problems

NASA Technical Reports Server (NTRS)

Koumoutsakos, P.

1995-01-01

We are developing computational tools for the simulations of three-dimensional flows past bodies undergoing arbitrary motions. High resolution viscous vortex methods have been developed that allow for extended simulations of two-dimensional configurations such as vortex generators. Our objective is to extend this methodology to three dimensions and develop a robust computational scheme for the simulation of such flows. A fundamental issue in the use of vortex methods is the ability of employing efficiently large numbers of computational elements to resolve the large range of scales that exist in complex flows. The traditional cost of the method scales as Omicron (N(sup 2)) as the N computational elements/particles induce velocities at each other, making the method unacceptable for simulations involving more than a few tens of thousands of particles. In the last decade fast methods have been developed that have operation counts of Omicron (N log N) or Omicron (N) (referred to as BH and GR respectively) depending on the details of the algorithm. These methods are based on the observation that the effect of a cluster of particles at a certain distance may be approximated by a finite series expansion. In order to exploit this observation we need to decompose the element population spatially into clusters of particles and build a hierarchy of clusters (a tree data structure) - smaller neighboring clusters combine to form a cluster of the next size up in the hierarchy and so on. This hierarchy of clusters allows one to determine efficiently when the approximation is valid. This algorithm is an N-body solver that appears in many fields of engineering and science. Some examples of its diverse use are in astrophysics, molecular dynamics, micro-magnetics, boundary element simulations of electromagnetic problems, and computer animation. More recently these N-body solvers have been implemented and applied in simulations involving vortex methods. Koumoutsakos and Leonard (1995) implemented the GR scheme in two dimensions for vector computer architectures allowing for simulations of bluff body flows using millions of particles. Winckelmans presented three-dimensional, viscous simulations of interacting vortex rings, using vortons and an implementation of a BH scheme for parallel computer architectures. Bhatt presented a vortex filament method to perform inviscid vortex ring interactions, with an alternative implementation of a BH scheme for a Connection Machine parallel computer architecture.

Importance of balanced architectures in the design of high-performance imaging systems

NASA Astrophysics Data System (ADS)

Sgro, Joseph A.; Stanton, Paul C.

1999-03-01

Imaging systems employed in demanding military and industrial applications, such as automatic target recognition and computer vision, typically require real-time high-performance computing resources. While high- performances computing systems have traditionally relied on proprietary architectures and custom components, recent advances in high performance general-purpose microprocessor technology have produced an abundance of low cost components suitable for use in high-performance computing systems. A common pitfall in the design of high performance imaging system, particularly systems employing scalable multiprocessor architectures, is the failure to balance computational and memory bandwidth. The performance of standard cluster designs, for example, in which several processors share a common memory bus, is typically constrained by memory bandwidth. The symptom characteristic of this problem is failure to the performance of the system to scale as more processors are added. The problem becomes exacerbated if I/O and memory functions share the same bus. The recent introduction of microprocessors with large internal caches and high performance external memory interfaces makes it practical to design high performance imaging system with balanced computational and memory bandwidth. Real word examples of such designs will be presented, along with a discussion of adapting algorithm design to best utilize available memory bandwidth.

Proposal for grid computing for nuclear applications

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

Idris, Faridah Mohamad; Ismail, Saaidi; Haris, Mohd Fauzi B.

2014-02-12

The use of computer clusters for computational sciences including computational physics is vital as it provides computing power to crunch big numbers at a faster rate. In compute intensive applications that requires high resolution such as Monte Carlo simulation, the use of computer clusters in a grid form that supplies computational power to any nodes within the grid that needs computing power, has now become a necessity. In this paper, we described how the clusters running on a specific application could use resources within the grid, to run the applications to speed up the computing process.

Efficient similarity-based data clustering by optimal object to cluster reallocation.

PubMed

Rossignol, Mathias; Lagrange, Mathieu; Cont, Arshia

2018-01-01

We present an iterative flat hard clustering algorithm designed to operate on arbitrary similarity matrices, with the only constraint that these matrices be symmetrical. Although functionally very close to kernel k-means, our proposal performs a maximization of average intra-class similarity, instead of a squared distance minimization, in order to remain closer to the semantics of similarities. We show that this approach permits the relaxing of some conditions on usable affinity matrices like semi-positiveness, as well as opening possibilities for computational optimization required for large datasets. Systematic evaluation on a variety of data sets shows that compared with kernel k-means and the spectral clustering methods, the proposed approach gives equivalent or better performance, while running much faster. Most notably, it significantly reduces memory access, which makes it a good choice for large data collections. Material enabling the reproducibility of the results is made available online.

Distributed Computing Architecture for Image-Based Wavefront Sensing and 2 D FFTs

NASA Technical Reports Server (NTRS)

Smith, Jeffrey S.; Dean, Bruce H.; Haghani, Shadan

2006-01-01

Image-based wavefront sensing (WFS) provides significant advantages over interferometric-based wavefi-ont sensors such as optical design simplicity and stability. However, the image-based approach is computational intensive, and therefore, specialized high-performance computing architectures are required in applications utilizing the image-based approach. The development and testing of these high-performance computing architectures are essential to such missions as James Webb Space Telescope (JWST), Terrestial Planet Finder-Coronagraph (TPF-C and CorSpec), and Spherical Primary Optical Telescope (SPOT). The development of these specialized computing architectures require numerous two-dimensional Fourier Transforms, which necessitate an all-to-all communication when applied on a distributed computational architecture. Several solutions for distributed computing are presented with an emphasis on a 64 Node cluster of DSPs, multiple DSP FPGAs, and an application of low-diameter graph theory. Timing results and performance analysis will be presented. The solutions offered could be applied to other all-to-all communication and scientifically computationally complex problems.

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

A Computational Cluster for Multiscale Simulations of Ionic Liquids

DTIC Science & Technology

2008-09-16

AND SUBTITLE DURIP: A Computational Cluster for Multiscale Simulations of Ionic Liquids 5a. CONTRACT NUMBER 5b. GRANT NUMBER FA955007-1-0512 5c...AVAILABILITY STATEMENT ZO\\5oc\\\\%1>^ 13. SUPPLEMENTARY NOTES 14. ABSTRACT The focus of this project was to acquire and use computer cluster nodes...by ANSI Std. Z39.18 Adobe Professional 7.0 Comprehensive Final Report: Gregory A. Voth, PI Contract/Grant Title: DURIP: A Computational Cluster for

Ab initio structures and polarizabilities of sodium clusters

NASA Astrophysics Data System (ADS)

Kronik, Leeor; Vasiliev, Igor; Jain, Manish; Chelikowsky, James R.

2001-09-01

We present quantitative ab initio calculations for Na cluster structures and polarizabilities, for all cluster sizes up to 20 atoms. Our calculations are performed by combining an ab initio core-corrected pseudopotential and a gradient-corrected density functional within a real space approach. We find the cluster bonding to be very floppy and catalog a host of low-energy quasi-degenerate isomers for all second-decade clusters. The existence of these isomers results in a band of polarizability values for each cluster size even at zero temperature. This eliminates any finer structure in the polarizability curve. We further show that the experimental polarizability values are consistently underestimated by calculations at zero temperature. By computing the effects of structure expansion and distortion due to a finite temperature we arrive at a quantitative agreement between theory and experiment.

Molgenis-impute: imputation pipeline in a box.

PubMed

Kanterakis, Alexandros; Deelen, Patrick; van Dijk, Freerk; Byelas, Heorhiy; Dijkstra, Martijn; Swertz, Morris A

2015-08-19

Genotype imputation is an important procedure in current genomic analysis such as genome-wide association studies, meta-analyses and fine mapping. Although high quality tools are available that perform the steps of this process, considerable effort and expertise is required to set up and run a best practice imputation pipeline, particularly for larger genotype datasets, where imputation has to scale out in parallel on computer clusters. Here we present MOLGENIS-impute, an 'imputation in a box' solution that seamlessly and transparently automates the set up and running of all the steps of the imputation process. These steps include genome build liftover (liftovering), genotype phasing with SHAPEIT2, quality control, sample and chromosomal chunking/merging, and imputation with IMPUTE2. MOLGENIS-impute builds on MOLGENIS-compute, a simple pipeline management platform for submission and monitoring of bioinformatics tasks in High Performance Computing (HPC) environments like local/cloud servers, clusters and grids. All the required tools, data and scripts are downloaded and installed in a single step. Researchers with diverse backgrounds and expertise have tested MOLGENIS-impute on different locations and imputed over 30,000 samples so far using the 1,000 Genomes Project and new Genome of the Netherlands data as the imputation reference. The tests have been performed on PBS/SGE clusters, cloud VMs and in a grid HPC environment. MOLGENIS-impute gives priority to the ease of setting up, configuring and running an imputation. It has minimal dependencies and wraps the pipeline in a simple command line interface, without sacrificing flexibility to adapt or limiting the options of underlying imputation tools. It does not require knowledge of a workflow system or programming, and is targeted at researchers who just want to apply best practices in imputation via simple commands. It is built on the MOLGENIS compute workflow framework to enable customization with additional computational steps or it can be included in other bioinformatics pipelines. It is available as open source from: https://github.com/molgenis/molgenis-imputation.

MRPack: Multi-Algorithm Execution Using Compute-Intensive Approach in MapReduce

PubMed Central

2015-01-01

Large quantities of data have been generated from multiple sources at exponential rates in the last few years. These data are generated at high velocity as real time and streaming data in variety of formats. These characteristics give rise to challenges in its modeling, computation, and processing. Hadoop MapReduce (MR) is a well known data-intensive distributed processing framework using the distributed file system (DFS) for Big Data. Current implementations of MR only support execution of a single algorithm in the entire Hadoop cluster. In this paper, we propose MapReducePack (MRPack), a variation of MR that supports execution of a set of related algorithms in a single MR job. We exploit the computational capability of a cluster by increasing the compute-intensiveness of MapReduce while maintaining its data-intensive approach. It uses the available computing resources by dynamically managing the task assignment and intermediate data. Intermediate data from multiple algorithms are managed using multi-key and skew mitigation strategies. The performance study of the proposed system shows that it is time, I/O, and memory efficient compared to the default MapReduce. The proposed approach reduces the execution time by 200% with an approximate 50% decrease in I/O cost. Complexity and qualitative results analysis shows significant performance improvement. PMID:26305223

MRPack: Multi-Algorithm Execution Using Compute-Intensive Approach in MapReduce.

PubMed

Idris, Muhammad; Hussain, Shujaat; Siddiqi, Muhammad Hameed; Hassan, Waseem; Syed Muhammad Bilal, Hafiz; Lee, Sungyoung

2015-01-01

Large quantities of data have been generated from multiple sources at exponential rates in the last few years. These data are generated at high velocity as real time and streaming data in variety of formats. These characteristics give rise to challenges in its modeling, computation, and processing. Hadoop MapReduce (MR) is a well known data-intensive distributed processing framework using the distributed file system (DFS) for Big Data. Current implementations of MR only support execution of a single algorithm in the entire Hadoop cluster. In this paper, we propose MapReducePack (MRPack), a variation of MR that supports execution of a set of related algorithms in a single MR job. We exploit the computational capability of a cluster by increasing the compute-intensiveness of MapReduce while maintaining its data-intensive approach. It uses the available computing resources by dynamically managing the task assignment and intermediate data. Intermediate data from multiple algorithms are managed using multi-key and skew mitigation strategies. The performance study of the proposed system shows that it is time, I/O, and memory efficient compared to the default MapReduce. The proposed approach reduces the execution time by 200% with an approximate 50% decrease in I/O cost. Complexity and qualitative results analysis shows significant performance improvement.

Ion induced electron emission statistics under Agm- cluster bombardment of Ag

NASA Astrophysics Data System (ADS)

Breuers, A.; Penning, R.; Wucher, A.

2018-05-01

The electron emission from a polycrystalline silver surface under bombardment with Agm- cluster ions (m = 1, 2, 3) is investigated in terms of ion induced kinetic excitation. The electron yield γ is determined directly by a current measurement method on the one hand and implicitly by the analysis of the electron emission statistics on the other hand. Successful measurements of the electron emission spectra ensure a deeper understanding of the ion induced kinetic electron emission process, with particular emphasis on the effect of the projectile cluster size to the yield as well as to emission statistics. The results allow a quantitative comparison to computer simulations performed for silver atoms and clusters impinging onto a silver surface.

Tetrahedral cluster and pseudo molecule: New approaches to Calculate Absolute Surface Energy of Zinc Blende (111)/(-1-1-1) Surface

NASA Astrophysics Data System (ADS)

Zhang, Yiou; Zhang, Jingzhao; Tse, Kinfai; Wong, Lun; Chan, Chunkai; Deng, Bei; Zhu, Junyi

Determining accurate absolute surface energies for polar surfaces of semiconductors has been a great challenge in decades. Here, we propose pseudo-hydrogen passivation to calculate them, using density functional theory approaches. By calculating the energy contribution from pseudo-hydrogen using either a pseudo molecule method or a tetrahedral cluster method, we obtained (111)/(-1-1-1) surfaces energies of Si, GaP, GaAs, and ZnS with high self-consistency. Our findings may greatly enhance the basic understandings of different surfaces and lead to novel strategies in the crystal growth. We would like to thank Su-huai Wei for helpful discussions. Computing resources were provided by the High Performance Cluster Computing Centre, Hong Kong Baptist University. This work was supported by the start-up funding and direct Grant with the Project.

Tracking the NGS revolution: managing life science research on shared high-performance computing clusters.

PubMed

Dahlö, Martin; Scofield, Douglas G; Schaal, Wesley; Spjuth, Ola

2018-05-01

Next-generation sequencing (NGS) has transformed the life sciences, and many research groups are newly dependent upon computer clusters to store and analyze large datasets. This creates challenges for e-infrastructures accustomed to hosting computationally mature research in other sciences. Using data gathered from our own clusters at UPPMAX computing center at Uppsala University, Sweden, where core hour usage of ∼800 NGS and ∼200 non-NGS projects is now similar, we compare and contrast the growth, administrative burden, and cluster usage of NGS projects with projects from other sciences. The number of NGS projects has grown rapidly since 2010, with growth driven by entry of new research groups. Storage used by NGS projects has grown more rapidly since 2013 and is now limited by disk capacity. NGS users submit nearly twice as many support tickets per user, and 11 more tools are installed each month for NGS projects than for non-NGS projects. We developed usage and efficiency metrics and show that computing jobs for NGS projects use more RAM than non-NGS projects, are more variable in core usage, and rarely span multiple nodes. NGS jobs use booked resources less efficiently for a variety of reasons. Active monitoring can improve this somewhat. Hosting NGS projects imposes a large administrative burden at UPPMAX due to large numbers of inexperienced users and diverse and rapidly evolving research areas. We provide a set of recommendations for e-infrastructures that host NGS research projects. We provide anonymized versions of our storage, job, and efficiency databases.

Tracking the NGS revolution: managing life science research on shared high-performance computing clusters

PubMed Central

2018-01-01

Abstract Background Next-generation sequencing (NGS) has transformed the life sciences, and many research groups are newly dependent upon computer clusters to store and analyze large datasets. This creates challenges for e-infrastructures accustomed to hosting computationally mature research in other sciences. Using data gathered from our own clusters at UPPMAX computing center at Uppsala University, Sweden, where core hour usage of ∼800 NGS and ∼200 non-NGS projects is now similar, we compare and contrast the growth, administrative burden, and cluster usage of NGS projects with projects from other sciences. Results The number of NGS projects has grown rapidly since 2010, with growth driven by entry of new research groups. Storage used by NGS projects has grown more rapidly since 2013 and is now limited by disk capacity. NGS users submit nearly twice as many support tickets per user, and 11 more tools are installed each month for NGS projects than for non-NGS projects. We developed usage and efficiency metrics and show that computing jobs for NGS projects use more RAM than non-NGS projects, are more variable in core usage, and rarely span multiple nodes. NGS jobs use booked resources less efficiently for a variety of reasons. Active monitoring can improve this somewhat. Conclusions Hosting NGS projects imposes a large administrative burden at UPPMAX due to large numbers of inexperienced users and diverse and rapidly evolving research areas. We provide a set of recommendations for e-infrastructures that host NGS research projects. We provide anonymized versions of our storage, job, and efficiency databases. PMID:29659792

Charliecloud

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

Priedhorsky, Reid; Randles, Tim

Charliecloud is a set of scripts to let users run a virtual cluster of virtual machines (VMs) on a desktop or supercomputer. Key functions include: 1. Creating (typically by installing an operating system from vendor media) and updating VM images; 2. Running a single VM; 3. Running multiple VMs in a virtual cluster. The virtual machines can talk to one another over the network and (in some cases) the outside world. This is accomplished by calling external programs such as QEMU and the Virtual Distributed Ethernet (VDE) suite. The goal is to let users have a virtual cluster containing nodesmore » where they have privileged access, while isolating that privilege within the virtual cluster so it cannot affect the physical compute resources. Host configuration enforces security; this is not included in Charliecloud, though security guidelines are included in its documentation and Charliecloud is designed to facilitate such configuration. Charliecloud manages passing information from host computers into and out of the virtual machines, such as parameters of the virtual cluster, input data specified by the user, output data from virtual compute jobs, VM console display, and network connections (e.g., SSH or X11). Parameters for the virtual cluster (number of VMs, RAM and disk per VM, etc.) are specified by the user or gathered from the environment (e.g., SLURM environment variables). Example job scripts are included. These include computation examples (such as a "hello world" MPI job) as well as performance tests. They also include a security test script to verify that the virtual cluster is appropriately sandboxed. Tests include: 1. Pinging hosts inside and outside the virtual cluster to explore connectivity; 2. Port scans (again inside and outside) to see what services are available; 3. Sniffing tests to see what traffic is visible to running VMs; 4. IP address spoofing to test network functionality in this case; 5. File access tests to make sure host access permissions are enforced. This test script is not a comprehensive scanner and does not test for specific vulnerabilities. Importantly, no information about physical hosts or network topology is included in this script (or any of Charliecloud); while part of a sensible test, such information is specified by the user when the test is run. That is, one cannot learn anything about the LANL network or computing infrastructure by examining Charliecloud code.« less

Cortico-hippocampal representations in simultaneous odor discrimination: a computational interpretation of Eichenbaum, Mathews, and Cohen (1989).

PubMed

Myers, C E; Gluck, M A

1996-08-01

A previous model of hippocampal region function in classical conditioning is generalized to H. Eichenbaum, A. Fagan, P. Mathews, and N.J. Cohen's (1989) and H. Eichenbaum, A. Fagan, and N.J. Cohen's (1989) simultaneous odor discrimination studies in rats. The model assumes that the hippocampal region forms new stimulus representations that compress redundant information while differentiating predictie information; the piriform (olfactory) cortex meanwhile clusters similar and co-occurring odors. Hippocampal damage interrupts the ability to differentiate odor representations, while leaving piriform-mediated odor clustering unchecked. The result is a net tendency to overcompress in the lesioned model. Behavior in the model is very similar to that of the rats, including lesion deficits, facilitation of successively learned tasks, and transfer performance. The computational mechanisms underlying model performance are consistent with the qualitative interpretations suggested by Eichen baum et al. to explain their empirical data.

Computer-aided classification of breast microcalcification clusters: merging of features from image processing and radiologists

NASA Astrophysics Data System (ADS)

Lo, Joseph Y.; Gavrielides, Marios A.; Markey, Mia K.; Jesneck, Jonathan L.

2003-05-01

We developed an ensemble classifier for the task of computer-aided diagnosis of breast microcalcification clusters,which are very challenging to characterize for radiologists and computer models alike. The purpose of this study is to help radiologists identify whether suspicious calcification clusters are benign vs. malignant, such that they may potentially recommend fewer unnecessary biopsies for actually benign lesions. The data consists of mammographic features extracted by automated image processing algorithms as well as manually interpreted by radiologists according to a standardized lexicon. We used 292 cases from a publicly available mammography database. From each cases, we extracted 22 image processing features pertaining to lesion morphology, 5 radiologist features also pertaining to morphology, and the patient age. Linear discriminant analysis (LDA) models were designed using each of the three data types. Each local model performed poorly; the best was one based upon image processing features which yielded ROC area index AZ of 0.59 +/- 0.03 and partial AZ above 90% sensitivity of 0.08 +/- 0.03. We then developed ensemble models using different combinations of those data types, and these models all improved performance compared to the local models. The final ensemble model was based upon 5 features selected by stepwise LDA from all 28 available features. This ensemble performed with AZ of 0.69 +/- 0.03 and partial AZ of 0.21 +/- 0.04, which was statistically significantly better than the model based on the image processing features alone (p<0.001 and p=0.01 for full and partial AZ respectively). This demonstrated the value of the radiologist-extracted features as a source of information for this task. It also suggested there is potential for improved performance using this ensemble classifier approach to combine different sources of currently available data.

Implementation of High-Order Multireference Coupled-Cluster Methods on Intel Many Integrated Core Architecture.

PubMed

Aprà, E; Kowalski, K

2016-03-08

In this paper we discuss the implementation of multireference coupled-cluster formalism with singles, doubles, and noniterative triples (MRCCSD(T)), which is capable of taking advantage of the processing power of the Intel Xeon Phi coprocessor. We discuss the integration of two levels of parallelism underlying the MRCCSD(T) implementation with computational kernels designed to offload the computationally intensive parts of the MRCCSD(T) formalism to Intel Xeon Phi coprocessors. Special attention is given to the enhancement of the parallel performance by task reordering that has improved load balancing in the noniterative part of the MRCCSD(T) calculations. We also discuss aspects regarding efficient optimization and vectorization strategies.

Response to "Comparison and Evaluation of Clustering Algorithms for Tandem Mass Spectra".

PubMed

Griss, Johannes; Perez-Riverol, Yasset; The, Matthew; Käll, Lukas; Vizcaíno, Juan Antonio

2018-05-04

In the recent benchmarking article entitled "Comparison and Evaluation of Clustering Algorithms for Tandem Mass Spectra", Rieder et al. compared several different approaches to cluster MS/MS spectra. While we certainly recognize the value of the manuscript, here, we report some shortcomings detected in the original analyses. For most analyses, the authors clustered only single MS/MS runs. In one of the reported analyses, three MS/MS runs were processed together, which already led to computational performance issues in many of the tested approaches. This fact highlights the difficulties of using many of the tested algorithms on the nowadays produced average proteomics data sets. Second, the authors only processed identified spectra when merging MS runs. Thereby, all unidentified spectra that are of lower quality were already removed from the data set and could not influence the clustering results. Next, we found that the authors did not analyze the effect of chimeric spectra on the clustering results. In our analysis, we found that 3% of the spectra in the used data sets were chimeric, and this had marked effects on the behavior of the different clustering algorithms tested. Finally, the authors' choice to evaluate the MS-Cluster and spectra-cluster algorithms using a precursor tolerance of 5 Da for high-resolution Orbitrap data only was, in our opinion, not adequate to assess the performance of MS/MS clustering approaches.

Oak Ridge Institutional Cluster Autotune Test Drive Report

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

Jibonananda, Sanyal; New, Joshua Ryan

2014-02-01

The Oak Ridge Institutional Cluster (OIC) provides general purpose computational resources for the ORNL staff to run computation heavy jobs that are larger than desktop applications but do not quite require the scale and power of the Oak Ridge Leadership Computing Facility (OLCF). This report details the efforts made and conclusions derived in performing a short test drive of the cluster resources on Phase 5 of the OIC. EnergyPlus was used in the analysis as a candidate user program and the overall software environment was evaluated against anticipated challenges experienced with resources such as the shared memory-Nautilus (JICS) and Titanmore » (OLCF). The OIC performed within reason and was found to be acceptable in the context of running EnergyPlus simulations. The number of cores per node and the availability of scratch space per node allow non-traditional desktop focused applications to leverage parallel ensemble execution. Although only individual runs of EnergyPlus were executed, the software environment on the OIC appeared suitable to run ensemble simulations with some modifications to the Autotune workflow. From a standpoint of general usability, the system supports common Linux libraries, compilers, standard job scheduling software (Torque/Moab), and the OpenMPI library (the only MPI library) for MPI communications. The file system is a Panasas file system which literature indicates to be an efficient file system.« less

Star clusters: age, metallicity and extinction from integrated spectra

NASA Astrophysics Data System (ADS)

González Delgado, Rosa M.; Cid Fernandes, Roberto

2010-01-01

Integrated optical spectra of star clusters in the Magellanic Clouds and a few Galactic globular clusters are fitted using high-resolution spectral models for single stellar populations. The goal is to estimate the age, metallicity and extinction of the clusters, and evaluate the degeneracies among these parameters. Several sets of evolutionary models that were computed with recent high-spectral-resolution stellar libraries (MILES, GRANADA, STELIB), are used as inputs to the starlight code to perform the fits. The comparison of the results derived from this method and previous estimates available in the literature allow us to evaluate the pros and cons of each set of models to determine star cluster properties. In addition, we quantify the uncertainties associated with the age, metallicity and extinction determinations resulting from variance in the ingredients for the analysis.

Analysis of basic clustering algorithms for numerical estimation of statistical averages in biomolecules.

PubMed

Anandakrishnan, Ramu; Onufriev, Alexey

2008-03-01

In statistical mechanics, the equilibrium properties of a physical system of particles can be calculated as the statistical average over accessible microstates of the system. In general, these calculations are computationally intractable since they involve summations over an exponentially large number of microstates. Clustering algorithms are one of the methods used to numerically approximate these sums. The most basic clustering algorithms first sub-divide the system into a set of smaller subsets (clusters). Then, interactions between particles within each cluster are treated exactly, while all interactions between different clusters are ignored. These smaller clusters have far fewer microstates, making the summation over these microstates, tractable. These algorithms have been previously used for biomolecular computations, but remain relatively unexplored in this context. Presented here, is a theoretical analysis of the error and computational complexity for the two most basic clustering algorithms that were previously applied in the context of biomolecular electrostatics. We derive a tight, computationally inexpensive, error bound for the equilibrium state of a particle computed via these clustering algorithms. For some practical applications, it is the root mean square error, which can be significantly lower than the error bound, that may be more important. We how that there is a strong empirical relationship between error bound and root mean square error, suggesting that the error bound could be used as a computationally inexpensive metric for predicting the accuracy of clustering algorithms for practical applications. An example of error analysis for such an application-computation of average charge of ionizable amino-acids in proteins-is given, demonstrating that the clustering algorithm can be accurate enough for practical purposes.

The coupling of fluids, dynamics, and controls on advanced architecture computers

NASA Technical Reports Server (NTRS)

Atwood, Christopher

1995-01-01

This grant provided for the demonstration of coupled controls, body dynamics, and fluids computations in a workstation cluster environment; and an investigation of the impact of peer-peer communication on flow solver performance and robustness. The findings of these investigations were documented in the conference articles.The attached publication, 'Towards Distributed Fluids/Controls Simulations', documents the solution and scaling of the coupled Navier-Stokes, Euler rigid-body dynamics, and state feedback control equations for a two-dimensional canard-wing. The poor scaling shown was due to serialized grid connectivity computation and Ethernet bandwidth limits. The scaling of a peer-to-peer communication flow code on an IBM SP-2 was also shown. The scaling of the code on the switched fabric-linked nodes was good, with a 2.4 percent loss due to communication of intergrid boundary point information. The code performance on 30 worker nodes was 1.7 (mu)s/point/iteration, or a factor of three over a Cray C-90 head. The attached paper, 'Nonlinear Fluid Computations in a Distributed Environment', documents the effect of several computational rate enhancing methods on convergence. For the cases shown, the highest throughput was achieved using boundary updates at each step, with the manager process performing communication tasks only. Constrained domain decomposition of the implicit fluid equations did not degrade the convergence rate or final solution. The scaling of a coupled body/fluid dynamics problem on an Ethernet-linked cluster was also shown.

Computational Science in Armenia (Invited Talk)

NASA Astrophysics Data System (ADS)

Marandjian, H.; Shoukourian, Yu.

This survey is devoted to the development of informatics and computer science in Armenia. The results in theoretical computer science (algebraic models, solutions to systems of general form recursive equations, the methods of coding theory, pattern recognition and image processing), constitute the theoretical basis for developing problem-solving-oriented environments. As examples can be mentioned: a synthesizer of optimized distributed recursive programs, software tools for cluster-oriented implementations of two-dimensional cellular automata, a grid-aware web interface with advanced service trading for linear algebra calculations. In the direction of solving scientific problems that require high-performance computing resources, examples of completed projects include the field of physics (parallel computing of complex quantum systems), astrophysics (Armenian virtual laboratory), biology (molecular dynamics study of human red blood cell membrane), meteorology (implementing and evaluating the Weather Research and Forecast Model for the territory of Armenia). The overview also notes that the Institute for Informatics and Automation Problems of the National Academy of Sciences of Armenia has established a scientific and educational infrastructure, uniting computing clusters of scientific and educational institutions of the country and provides the scientific community with access to local and international computational resources, that is a strong support for computational science in Armenia.

A Decentralized Eigenvalue Computation Method for Spectrum Sensing Based on Average Consensus

NASA Astrophysics Data System (ADS)

Mohammadi, Jafar; Limmer, Steffen; Stańczak, Sławomir

2016-07-01

This paper considers eigenvalue estimation for the decentralized inference problem for spectrum sensing. We propose a decentralized eigenvalue computation algorithm based on the power method, which is referred to as generalized power method GPM; it is capable of estimating the eigenvalues of a given covariance matrix under certain conditions. Furthermore, we have developed a decentralized implementation of GPM by splitting the iterative operations into local and global computation tasks. The global tasks require data exchange to be performed among the nodes. For this task, we apply an average consensus algorithm to efficiently perform the global computations. As a special case, we consider a structured graph that is a tree with clusters of nodes at its leaves. For an accelerated distributed implementation, we propose to use computation over multiple access channel (CoMAC) as a building block of the algorithm. Numerical simulations are provided to illustrate the performance of the two algorithms.

Experiences modeling ocean circulation problems on a 30 node commodity cluster with 3840 GPU processor cores.

NASA Astrophysics Data System (ADS)

Hill, C.

2008-12-01

Low cost graphic cards today use many, relatively simple, compute cores to deliver support for memory bandwidth of more than 100GB/s and theoretical floating point performance of more than 500 GFlop/s. Right now this performance is, however, only accessible to highly parallel algorithm implementations that, (i) can use a hundred or more, 32-bit floating point, concurrently executing cores, (ii) can work with graphics memory that resides on the graphics card side of the graphics bus and (iii) can be partially expressed in a language that can be compiled by a graphics programming tool. In this talk we describe our experiences implementing a complete, but relatively simple, time dependent shallow-water equations simulation targeting a cluster of 30 computers each hosting one graphics card. The implementation takes into account the considerations (i), (ii) and (iii) listed previously. We code our algorithm as a series of numerical kernels. Each kernel is designed to be executed by multiple threads of a single process. Kernels are passed memory blocks to compute over which can be persistent blocks of memory on a graphics card. Each kernel is individually implemented using the NVidia CUDA language but driven from a higher level supervisory code that is almost identical to a standard model driver. The supervisory code controls the overall simulation timestepping, but is written to minimize data transfer between main memory and graphics memory (a massive performance bottle-neck on current systems). Using the recipe outlined we can boost the performance of our cluster by nearly an order of magnitude, relative to the same algorithm executing only on the cluster CPU's. Achieving this performance boost requires that many threads are available to each graphics processor for execution within each numerical kernel and that the simulations working set of data can fit into the graphics card memory. As we describe, this puts interesting upper and lower bounds on the problem sizes for which this technology is currently most useful. However, many interesting problems fit within this envelope. Looking forward, we extrapolate our experience to estimate full-scale ocean model performance and applicability. Finally we describe preliminary hybrid mixed 32-bit and 64-bit experiments with graphics cards that support 64-bit arithmetic, albeit at a lower performance.

Cluster Computing for Embedded/Real-Time Systems

NASA Technical Reports Server (NTRS)

Katz, D.; Kepner, J.

1999-01-01

Embedded and real-time systems, like other computing systems, seek to maximize computing power for a given price, and thus can significantly benefit from the advancing capabilities of cluster computing.

Coarse-Grained Clustering Dynamics of Heterogeneously Coupled Neurons.

PubMed

Moon, Sung Joon; Cook, Katherine A; Rajendran, Karthikeyan; Kevrekidis, Ioannis G; Cisternas, Jaime; Laing, Carlo R

2015-12-01

The formation of oscillating phase clusters in a network of identical Hodgkin-Huxley neurons is studied, along with their dynamic behavior. The neurons are synaptically coupled in an all-to-all manner, yet the synaptic coupling characteristic time is heterogeneous across the connections. In a network of N neurons where this heterogeneity is characterized by a prescribed random variable, the oscillatory single-cluster state can transition-through [Formula: see text] (possibly perturbed) period-doubling and subsequent bifurcations-to a variety of multiple-cluster states. The clustering dynamic behavior is computationally studied both at the detailed and the coarse-grained levels, and a numerical approach that can enable studying the coarse-grained dynamics in a network of arbitrarily large size is suggested. Among a number of cluster states formed, double clusters, composed of nearly equal sub-network sizes are seen to be stable; interestingly, the heterogeneity parameter in each of the double-cluster components tends to be consistent with the random variable over the entire network: Given a double-cluster state, permuting the dynamical variables of the neurons can lead to a combinatorially large number of different, yet similar "fine" states that appear practically identical at the coarse-grained level. For weak heterogeneity we find that correlations rapidly develop, within each cluster, between the neuron's "identity" (its own value of the heterogeneity parameter) and its dynamical state. For single- and double-cluster states we demonstrate an effective coarse-graining approach that uses the Polynomial Chaos expansion to succinctly describe the dynamics by these quickly established "identity-state" correlations. This coarse-graining approach is utilized, within the equation-free framework, to perform efficient computations of the neuron ensemble dynamics.

Running VisIt Software on the Peregrine System | High-Performance Computing

Science.gov Websites

kilobyte range. VisIt features a robust remote visualization capability. VisIt can be started on a local machine and used to visualize data on a remote compute cluster.The remote machine must be able to send VisIt module must be loaded as part of this process. To enable remote visualization the 'module load

Translational bioinformatics in the cloud: an affordable alternative

PubMed Central

2010-01-01

With the continued exponential expansion of publicly available genomic data and access to low-cost, high-throughput molecular technologies for profiling patient populations, computational technologies and informatics are becoming vital considerations in genomic medicine. Although cloud computing technology is being heralded as a key enabling technology for the future of genomic research, available case studies are limited to applications in the domain of high-throughput sequence data analysis. The goal of this study was to evaluate the computational and economic characteristics of cloud computing in performing a large-scale data integration and analysis representative of research problems in genomic medicine. We find that the cloud-based analysis compares favorably in both performance and cost in comparison to a local computational cluster, suggesting that cloud computing technologies might be a viable resource for facilitating large-scale translational research in genomic medicine. PMID:20691073

Shared Memory Parallelization of an Implicit ADI-type CFD Code

NASA Technical Reports Server (NTRS)

Hauser, Th.; Huang, P. G.

1999-01-01

A parallelization study designed for ADI-type algorithms is presented using the OpenMP specification for shared-memory multiprocessor programming. Details of optimizations specifically addressed to cache-based computer architectures are described and performance measurements for the single and multiprocessor implementation are summarized. The paper demonstrates that optimization of memory access on a cache-based computer architecture controls the performance of the computational algorithm. A hybrid MPI/OpenMP approach is proposed for clusters of shared memory machines to further enhance the parallel performance. The method is applied to develop a new LES/DNS code, named LESTool. A preliminary DNS calculation of a fully developed channel flow at a Reynolds number of 180, Re(sub tau) = 180, has shown good agreement with existing data.

Supermassive Black Hole Binaries in High Performance Massively Parallel Direct N-body Simulations on Large GPU Clusters

NASA Astrophysics Data System (ADS)

Spurzem, R.; Berczik, P.; Zhong, S.; Nitadori, K.; Hamada, T.; Berentzen, I.; Veles, A.

2012-07-01

Astrophysical Computer Simulations of Dense Star Clusters in Galactic Nuclei with Supermassive Black Holes are presented using new cost-efficient supercomputers in China accelerated by graphical processing cards (GPU). We use large high-accuracy direct N-body simulations with Hermite scheme and block-time steps, parallelised across a large number of nodes on the large scale and across many GPU thread processors on each node on the small scale. A sustained performance of more than 350 Tflop/s for a science run on using simultaneously 1600 Fermi C2050 GPUs is reached; a detailed performance model is presented and studies for the largest GPU clusters in China with up to Petaflop/s performance and 7000 Fermi GPU cards. In our case study we look at two supermassive black holes with equal and unequal masses embedded in a dense stellar cluster in a galactic nucleus. The hardening processes due to interactions between black holes and stars, effects of rotation in the stellar system and relativistic forces between the black holes are simultaneously taken into account. The simulation stops at the complete relativistic merger of the black holes.

The performance of low-cost commercial cloud computing as an alternative in computational chemistry.

PubMed

Thackston, Russell; Fortenberry, Ryan C

2015-05-05

The growth of commercial cloud computing (CCC) as a viable means of computational infrastructure is largely unexplored for the purposes of quantum chemistry. In this work, the PSI4 suite of computational chemistry programs is installed on five different types of Amazon World Services CCC platforms. The performance for a set of electronically excited state single-point energies is compared between these CCC platforms and typical, "in-house" physical machines. Further considerations are made for the number of cores or virtual CPUs (vCPUs, for the CCC platforms), but no considerations are made for full parallelization of the program (even though parallelization of the BLAS library is implemented), complete high-performance computing cluster utilization, or steal time. Even with this most pessimistic view of the computations, CCC resources are shown to be more cost effective for significant numbers of typical quantum chemistry computations. Large numbers of large computations are still best utilized by more traditional means, but smaller-scale research may be more effectively undertaken through CCC services. © 2015 Wiley Periodicals, Inc.

Screen media usage, sleep time and academic performance in adolescents: clustering a self-organizing maps analysis.

PubMed

Peiró-Velert, Carmen; Valencia-Peris, Alexandra; González, Luis M; García-Massó, Xavier; Serra-Añó, Pilar; Devís-Devís, José

2014-01-01

Screen media usage, sleep time and socio-demographic features are related to adolescents' academic performance, but interrelations are little explored. This paper describes these interrelations and behavioral profiles clustered in low and high academic performance. A nationally representative sample of 3,095 Spanish adolescents, aged 12 to 18, was surveyed on 15 variables linked to the purpose of the study. A Self-Organizing Maps analysis established non-linear interrelationships among these variables and identified behavior patterns in subsequent cluster analyses. Topological interrelationships established from the 15 emerging maps indicated that boys used more passive videogames and computers for playing than girls, who tended to use mobile phones to communicate with others. Adolescents with the highest academic performance were the youngest. They slept more and spent less time using sedentary screen media when compared to those with the lowest performance, and they also showed topological relationships with higher socioeconomic status adolescents. Cluster 1 grouped boys who spent more than 5.5 hours daily using sedentary screen media. Their academic performance was low and they slept an average of 8 hours daily. Cluster 2 gathered girls with an excellent academic performance, who slept nearly 9 hours per day, and devoted less time daily to sedentary screen media. Academic performance was directly related to sleep time and socioeconomic status, but inversely related to overall sedentary screen media usage. Profiles from the two clusters were strongly differentiated by gender, age, sedentary screen media usage, sleep time and academic achievement. Girls with the highest academic results had a medium socioeconomic status in Cluster 2. Findings may contribute to establishing recommendations about the timing and duration of screen media usage in adolescents and appropriate sleep time needed to successfully meet the demands of school academics and to improve interventions targeting to affect behavioral change.

Screen Media Usage, Sleep Time and Academic Performance in Adolescents: Clustering a Self-Organizing Maps Analysis

PubMed Central

Peiró-Velert, Carmen; Valencia-Peris, Alexandra; González, Luis M.; García-Massó, Xavier; Serra-Añó, Pilar; Devís-Devís, José

2014-01-01

Screen media usage, sleep time and socio-demographic features are related to adolescents' academic performance, but interrelations are little explored. This paper describes these interrelations and behavioral profiles clustered in low and high academic performance. A nationally representative sample of 3,095 Spanish adolescents, aged 12 to 18, was surveyed on 15 variables linked to the purpose of the study. A Self-Organizing Maps analysis established non-linear interrelationships among these variables and identified behavior patterns in subsequent cluster analyses. Topological interrelationships established from the 15 emerging maps indicated that boys used more passive videogames and computers for playing than girls, who tended to use mobile phones to communicate with others. Adolescents with the highest academic performance were the youngest. They slept more and spent less time using sedentary screen media when compared to those with the lowest performance, and they also showed topological relationships with higher socioeconomic status adolescents. Cluster 1 grouped boys who spent more than 5.5 hours daily using sedentary screen media. Their academic performance was low and they slept an average of 8 hours daily. Cluster 2 gathered girls with an excellent academic performance, who slept nearly 9 hours per day, and devoted less time daily to sedentary screen media. Academic performance was directly related to sleep time and socioeconomic status, but inversely related to overall sedentary screen media usage. Profiles from the two clusters were strongly differentiated by gender, age, sedentary screen media usage, sleep time and academic achievement. Girls with the highest academic results had a medium socioeconomic status in Cluster 2. Findings may contribute to establishing recommendations about the timing and duration of screen media usage in adolescents and appropriate sleep time needed to successfully meet the demands of school academics and to improve interventions targeting to affect behavioral change. PMID:24941009

Construction and application of Red5 cluster based on OpenStack

NASA Astrophysics Data System (ADS)

Wang, Jiaqing; Song, Jianxin

2017-08-01

With the application and development of cloud computing technology in various fields, the resource utilization rate of the data center has been improved obviously, and the system based on cloud computing platform has also improved the expansibility and stability. In the traditional way, Red5 cluster resource utilization is low and the system stability is poor. This paper uses cloud computing to efficiently calculate the resource allocation ability, and builds a Red5 server cluster based on OpenStack. Multimedia applications can be published to the Red5 cloud server cluster. The system achieves the flexible construction of computing resources, but also greatly improves the stability of the cluster and service efficiency.

Application of microarray analysis on computer cluster and cloud platforms.

PubMed

Bernau, C; Boulesteix, A-L; Knaus, J

2013-01-01

Analysis of recent high-dimensional biological data tends to be computationally intensive as many common approaches such as resampling or permutation tests require the basic statistical analysis to be repeated many times. A crucial advantage of these methods is that they can be easily parallelized due to the computational independence of the resampling or permutation iterations, which has induced many statistics departments to establish their own computer clusters. An alternative is to rent computing resources in the cloud, e.g. at Amazon Web Services. In this article we analyze whether a selection of statistical projects, recently implemented at our department, can be efficiently realized on these cloud resources. Moreover, we illustrate an opportunity to combine computer cluster and cloud resources. In order to compare the efficiency of computer cluster and cloud implementations and their respective parallelizations we use microarray analysis procedures and compare their runtimes on the different platforms. Amazon Web Services provide various instance types which meet the particular needs of the different statistical projects we analyzed in this paper. Moreover, the network capacity is sufficient and the parallelization is comparable in efficiency to standard computer cluster implementations. Our results suggest that many statistical projects can be efficiently realized on cloud resources. It is important to mention, however, that workflows can change substantially as a result of a shift from computer cluster to cloud computing.

A Hardware-Accelerated Quantum Monte Carlo framework (HAQMC) for N-body systems

NASA Astrophysics Data System (ADS)

Gothandaraman, Akila; Peterson, Gregory D.; Warren, G. Lee; Hinde, Robert J.; Harrison, Robert J.

2009-12-01

Interest in the study of structural and energetic properties of highly quantum clusters, such as inert gas clusters has motivated the development of a hardware-accelerated framework for Quantum Monte Carlo simulations. In the Quantum Monte Carlo method, the properties of a system of atoms, such as the ground-state energies, are averaged over a number of iterations. Our framework is aimed at accelerating the computations in each iteration of the QMC application by offloading the calculation of properties, namely energy and trial wave function, onto reconfigurable hardware. This gives a user the capability to run simulations for a large number of iterations, thereby reducing the statistical uncertainty in the properties, and for larger clusters. This framework is designed to run on the Cray XD1 high performance reconfigurable computing platform, which exploits the coarse-grained parallelism of the processor along with the fine-grained parallelism of the reconfigurable computing devices available in the form of field-programmable gate arrays. In this paper, we illustrate the functioning of the framework, which can be used to calculate the energies for a model cluster of helium atoms. In addition, we present the capabilities of the framework that allow the user to vary the chemical identities of the simulated atoms. Program summaryProgram title: Hardware Accelerated Quantum Monte Carlo (HAQMC) Catalogue identifier: AEEP_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEEP_v1_0.html Program obtainable from: CPC Program Library, Queen's University, Belfast, N. Ireland Licensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.html No. of lines in distributed program, including test data, etc.: 691 537 No. of bytes in distributed program, including test data, etc.: 5 031 226 Distribution format: tar.gz Programming language: C/C++ for the QMC application, VHDL and Xilinx 8.1 ISE/EDK tools for FPGA design and development Computer: Cray XD1 consisting of a dual-core, dualprocessor AMD Opteron 2.2 GHz with a Xilinx Virtex-4 (V4LX160) or Xilinx Virtex-II Pro (XC2VP50) FPGA per node. We use the compute node with the Xilinx Virtex-4 FPGA Operating system: Red Hat Enterprise Linux OS Has the code been vectorised or parallelized?: Yes Classification: 6.1 Nature of problem: Quantum Monte Carlo is a practical method to solve the Schrödinger equation for large many-body systems and obtain the ground-state properties of such systems. This method involves the sampling of a number of configurations of atoms and averaging the properties of the configurations over a number of iterations. We are interested in applying the QMC method to obtain the energy and other properties of highly quantum clusters, such as inert gas clusters. Solution method: The proposed framework provides a combined hardware-software approach, in which the QMC simulation is performed on the host processor, with the computationally intensive functions such as energy and trial wave function computations mapped onto the field-programmable gate array (FPGA) logic device attached as a co-processor to the host processor. We perform the QMC simulation for a number of iterations as in the case of our original software QMC approach, to reduce the statistical uncertainty of the results. However, our proposed HAQMC framework accelerates each iteration of the simulation, by significantly reducing the time taken to calculate the ground-state properties of the configurations of atoms, thereby accelerating the overall QMC simulation. We provide a generic interpolation framework that can be extended to study a variety of pure and doped atomic clusters, irrespective of the chemical identities of the atoms. For the FPGA implementation of the properties, we use a two-region approach for accurately computing the properties over the entire domain, employ deep pipelines and fixed-point for all our calculations guaranteeing the accuracy required for our simulation.

High Performance Computing Based Parallel HIearchical Modal Association Clustering (HPAR HMAC)

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

Patlolla, Dilip R; Surendran Nair, Sujithkumar; Graves, Daniel A.

For many applications, clustering is a crucial step in order to gain insight into the makeup of a dataset. The best approach to a given problem often depends on a variety of factors, such as the size of the dataset, time restrictions, and soft clustering requirements. The HMAC algorithm seeks to combine the strengths of 2 particular clustering approaches: model-based and linkage-based clustering. One particular weakness of HMAC is its computational complexity. HMAC is not practical for mega-scale data clustering. For high-definition imagery, a user would have to wait months or years for a result; for a 16-megapixel image, themore » estimated runtime skyrockets to over a decade! To improve the execution time of HMAC, it is reasonable to consider an multi-core implementation that utilizes available system resources. An existing imple-mentation (Ray and Cheng 2014) divides the dataset into N partitions - one for each thread prior to executing the HMAC algorithm. This implementation benefits from 2 types of optimization: parallelization and divide-and-conquer. By running each partition in parallel, the program is able to accelerate computation by utilizing more system resources. Although the parallel implementation provides considerable improvement over the serial HMAC, it still suffers from poor computational complexity, O(N2). Once the maximum number of cores on a system is exhausted, the program exhibits slower behavior. We now consider a modification to HMAC that involves a recursive partitioning scheme. Our modification aims to exploit divide-and-conquer benefits seen by the parallel HMAC implementation. At each level in the recursion tree, partitions are divided into 2 sub-partitions until a threshold size is reached. When the partition can no longer be divided without falling below threshold size, the base HMAC algorithm is applied. This results in a significant speedup over the parallel HMAC.« less

Efficient Agent-Based Cluster Ensembles

NASA Technical Reports Server (NTRS)

Agogino, Adrian; Tumer, Kagan

2006-01-01

Numerous domains ranging from distributed data acquisition to knowledge reuse need to solve the cluster ensemble problem of combining multiple clusterings into a single unified clustering. Unfortunately current non-agent-based cluster combining methods do not work in a distributed environment, are not robust to corrupted clusterings and require centralized access to all original clusterings. Overcoming these issues will allow cluster ensembles to be used in fundamentally distributed and failure-prone domains such as data acquisition from satellite constellations, in addition to domains demanding confidentiality such as combining clusterings of user profiles. This paper proposes an efficient, distributed, agent-based clustering ensemble method that addresses these issues. In this approach each agent is assigned a small subset of the data and votes on which final cluster its data points should belong to. The final clustering is then evaluated by a global utility, computed in a distributed way. This clustering is also evaluated using an agent-specific utility that is shown to be easier for the agents to maximize. Results show that agents using the agent-specific utility can achieve better performance than traditional non-agent based methods and are effective even when up to 50% of the agents fail.

Clustering evolving proteins into homologous families.

PubMed

Chan, Cheong Xin; Mahbob, Maisarah; Ragan, Mark A

2013-04-08

Clustering sequences into groups of putative homologs (families) is a critical first step in many areas of comparative biology and bioinformatics. The performance of clustering approaches in delineating biologically meaningful families depends strongly on characteristics of the data, including content bias and degree of divergence. New, highly scalable methods have recently been introduced to cluster the very large datasets being generated by next-generation sequencing technologies. However, there has been little systematic investigation of how characteristics of the data impact the performance of these approaches. Using clusters from a manually curated dataset as reference, we examined the performance of a widely used graph-based Markov clustering algorithm (MCL) and a greedy heuristic approach (UCLUST) in delineating protein families coded by three sets of bacterial genomes of different G+C content. Both MCL and UCLUST generated clusters that are comparable to the reference sets at specific parameter settings, although UCLUST tends to under-cluster compositionally biased sequences (G+C content 33% and 66%). Using simulated data, we sought to assess the individual effects of sequence divergence, rate heterogeneity, and underlying G+C content. Performance decreased with increasing sequence divergence, decreasing among-site rate variation, and increasing G+C bias. Two MCL-based methods recovered the simulated families more accurately than did UCLUST. MCL using local alignment distances is more robust across the investigated range of sequence features than are greedy heuristics using distances based on global alignment. Our results demonstrate that sequence divergence, rate heterogeneity and content bias can individually and in combination affect the accuracy with which MCL and UCLUST can recover homologous protein families. For application to data that are more divergent, and exhibit higher among-site rate variation and/or content bias, MCL may often be the better choice, especially if computational resources are not limiting.

A Computational Linguistic Measure of Clustering Behavior on Semantic Verbal Fluency Task Predicts Risk of Future Dementia in the Nun Study

PubMed Central

Pakhomov, Serguei V.S.; Hemmy, Laura S.

2014-01-01

Generative semantic verbal fluency (SVF) tests show early and disproportionate decline relative to other abilities in individuals developing Alzheimer’s disease. Optimal performance on SVF tests depends on the efficiency of using clustered organization of semantically related items and the ability to switch between clusters. Traditional approaches to clustering and switching have relied on manual determination of clusters. We evaluated a novel automated computational linguistic approach for quantifying clustering behavior. Our approach is based on Latent Semantic Analysis (LSA) for computing strength of semantic relatedness between pairs of words produced in response to SVF test. The mean size of semantic clusters (MCS) and semantic chains (MChS) are calculated based on pairwise relatedness values between words. We evaluated the predictive validity of these measures on a set of 239 participants in the Nun Study, a longitudinal study of aging. All were cognitively intact at baseline assessment, measured with the CERAD battery, and were followed in 18 month waves for up to 20 years. The onset of either dementia or memory impairment were used as outcomes in Cox proportional hazards models adjusted for age and education and censored at follow up waves 5 (6.3 years) and 13 (16.96 years). Higher MCS was associated with 38% reduction in dementia risk at wave 5 and 26% reduction at wave 13, but not with the onset of memory impairment. Higher (+1 SD) MChS was associated with 39% dementia risk reduction at wave 5 but not wave 13, and association with memory impairment was not significant. Higher traditional SVF scores were associated with 22–29% memory impairment and 35–40% dementia risk reduction. SVF scores were not correlated with either MCS or MChS. Our study suggests that an automated approach to measuring clustering behavior can be used to estimate dementia risk in cognitively normal individuals. PMID:23845236

A computational linguistic measure of clustering behavior on semantic verbal fluency task predicts risk of future dementia in the nun study.

PubMed

Pakhomov, Serguei V S; Hemmy, Laura S

2014-06-01

Generative semantic verbal fluency (SVF) tests show early and disproportionate decline relative to other abilities in individuals developing Alzheimer's disease. Optimal performance on SVF tests depends on the efficiency of using clustered organization of semantically related items and the ability to switch between clusters. Traditional approaches to clustering and switching have relied on manual determination of clusters. We evaluated a novel automated computational linguistic approach for quantifying clustering behavior. Our approach is based on Latent Semantic Analysis (LSA) for computing strength of semantic relatedness between pairs of words produced in response to SVF test. The mean size of semantic clusters (MCS) and semantic chains (MChS) are calculated based on pairwise relatedness values between words. We evaluated the predictive validity of these measures on a set of 239 participants in the Nun Study, a longitudinal study of aging. All were cognitively intact at baseline assessment, measured with the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) battery, and were followed in 18-month waves for up to 20 years. The onset of either dementia or memory impairment were used as outcomes in Cox proportional hazards models adjusted for age and education and censored at follow-up waves 5 (6.3 years) and 13 (16.96 years). Higher MCS was associated with 38% reduction in dementia risk at wave 5 and 26% reduction at wave 13, but not with the onset of memory impairment. Higher [+1 standard deviation (SD)] MChS was associated with 39% dementia risk reduction at wave 5 but not wave 13, and association with memory impairment was not significant. Higher traditional SVF scores were associated with 22-29% memory impairment and 35-40% dementia risk reduction. SVF scores were not correlated with either MCS or MChS. Our study suggests that an automated approach to measuring clustering behavior can be used to estimate dementia risk in cognitively normal individuals. Copyright © 2013 Elsevier Ltd. All rights reserved.

The Roots of Beowulf

NASA Technical Reports Server (NTRS)

Fischer, James R.

2014-01-01

The first Beowulf Linux commodity cluster was constructed at NASA's Goddard Space Flight Center in 1994 and its origins are a part of the folklore of high-end computing. In fact, the conditions within Goddard that brought the idea into being were shaped by rich historical roots, strategic pressures brought on by the ramp up of the Federal High-Performance Computing and Communications Program, growth of the open software movement, microprocessor performance trends, and the vision of key technologists. This multifaceted story is told here for the first time from the point of view of NASA project management.

A multiple-feature and multiple-kernel scene segmentation algorithm for humanoid robot.

PubMed

Liu, Zhi; Xu, Shuqiong; Zhang, Yun; Chen, Chun Lung Philip

2014-11-01

This technical correspondence presents a multiple-feature and multiple-kernel support vector machine (MFMK-SVM) methodology to achieve a more reliable and robust segmentation performance for humanoid robot. The pixel wise intensity, gradient, and C1 SMF features are extracted via the local homogeneity model and Gabor filter, which would be used as inputs of MFMK-SVM model. It may provide multiple features of the samples for easier implementation and efficient computation of MFMK-SVM model. A new clustering method, which is called feature validity-interval type-2 fuzzy C-means (FV-IT2FCM) clustering algorithm, is proposed by integrating a type-2 fuzzy criterion in the clustering optimization process to improve the robustness and reliability of clustering results by the iterative optimization. Furthermore, the clustering validity is employed to select the training samples for the learning of the MFMK-SVM model. The MFMK-SVM scene segmentation method is able to fully take advantage of the multiple features of scene image and the ability of multiple kernels. Experiments on the BSDS dataset and real natural scene images demonstrate the superior performance of our proposed method.

Efficient Deployment of Key Nodes for Optimal Coverage of Industrial Mobile Wireless Networks

PubMed Central

Li, Xiaomin; Li, Di; Dong, Zhijie; Hu, Yage; Liu, Chengliang

2018-01-01

In recent years, industrial wireless networks (IWNs) have been transformed by the introduction of mobile nodes, and they now offer increased extensibility, mobility, and flexibility. Nevertheless, mobile nodes pose efficiency and reliability challenges. Efficient node deployment and management of channel interference directly affect network system performance, particularly for key node placement in clustered wireless networks. This study analyzes this system model, considering both industrial properties of wireless networks and their mobility. Then, static and mobile node coverage problems are unified and simplified to target coverage problems. We propose a novel strategy for the deployment of clustered heads in grouped industrial mobile wireless networks (IMWNs) based on the improved maximal clique model and the iterative computation of new candidate cluster head positions. The maximal cliques are obtained via a double-layer Tabu search. Each cluster head updates its new position via an improved virtual force while moving with full coverage to find the minimal inter-cluster interference. Finally, we develop a simulation environment. The simulation results, based on a performance comparison, show the efficacy of the proposed strategies and their superiority over current approaches. PMID:29439439

Graph-Based Object Class Discovery

NASA Astrophysics Data System (ADS)

Xia, Shengping; Hancock, Edwin R.

We are interested in the problem of discovering the set of object classes present in a database of images using a weakly supervised graph-based framework. Rather than making use of the ”Bag-of-Features (BoF)” approach widely used in current work on object recognition, we represent each image by a graph using a group of selected local invariant features. Using local feature matching and iterative Procrustes alignment, we perform graph matching and compute a similarity measure. Borrowing the idea of query expansion , we develop a similarity propagation based graph clustering (SPGC) method. Using this method class specific clusters of the graphs can be obtained. Such a cluster can be generally represented by using a higher level graph model whose vertices are the clustered graphs, and the edge weights are determined by the pairwise similarity measure. Experiments are performed on a dataset, in which the number of images increases from 1 to 50K and the number of objects increases from 1 to over 500. Some objects have been discovered with total recall and a precision 1 in a single cluster.

Quantum Dynamics of Helium Clusters

DTIC Science & Technology

1993-03-01

the structure of both these and the HeN clusters in the body fixed frame by computing principal moments of inertia, thereby avoiding the...8217 of helium clusters, with the modification that we subtract 0.96 K from the computed values so that lor sufficiently large clusters we recover the...phonon spectrum of liquid He. To get a picture of these spectra one needs to compute the structure functions 51. Monte Carlo random walk simulations

Cots Correlator Platform

NASA Astrophysics Data System (ADS)

Schaaf, Kjeld; Overeem, Ruud

2004-06-01

Moore’s law is best exploited by using consumer market hardware. In particular, the gaming industry pushes the limit of processor performance thus reducing the cost per raw flop even faster than Moore’s law predicts. Next to the cost benefits of Common-Of-The-Shelf (COTS) processing resources, there is a rapidly growing experience pool in cluster based processing. The typical Beowulf cluster of PC’s supercomputers are well known. Multiple examples exists of specialised cluster computers based on more advanced server nodes or even gaming stations. All these cluster machines build upon the same knowledge about cluster software management, scheduling, middleware libraries and mathematical libraries. In this study, we have integrated COTS processing resources and cluster nodes into a very high performance processing platform suitable for streaming data applications, in particular to implement a correlator. The required processing power for the correlator in modern radio telescopes is in the range of the larger supercomputers, which motivates the usage of supercomputer technology. Raw processing power is provided by graphical processors and is combined with an Infiniband host bus adapter with integrated data stream handling logic. With this processing platform a scalable correlator can be built with continuously growing processing power at consumer market prices.

Grid heterogeneity in in-silico experiments: an exploration of drug screening using DOCK on cloud environments.

PubMed

Yim, Wen-Wai; Chien, Shu; Kusumoto, Yasuyuki; Date, Susumu; Haga, Jason

2010-01-01

Large-scale in-silico screening is a necessary part of drug discovery and Grid computing is one answer to this demand. A disadvantage of using Grid computing is the heterogeneous computational environments characteristic of a Grid. In our study, we have found that for the molecular docking simulation program DOCK, different clusters within a Grid organization can yield inconsistent results. Because DOCK in-silico virtual screening (VS) is currently used to help select chemical compounds to test with in-vitro experiments, such differences have little effect on the validity of using virtual screening before subsequent steps in the drug discovery process. However, it is difficult to predict whether the accumulation of these discrepancies over sequentially repeated VS experiments will significantly alter the results if VS is used as the primary means for identifying potential drugs. Moreover, such discrepancies may be unacceptable for other applications requiring more stringent thresholds. This highlights the need for establishing a more complete solution to provide the best scientific accuracy when executing an application across Grids. One possible solution to platform heterogeneity in DOCK performance explored in our study involved the use of virtual machines as a layer of abstraction. This study investigated the feasibility and practicality of using virtual machine and recent cloud computing technologies in a biological research application. We examined the differences and variations of DOCK VS variables, across a Grid environment composed of different clusters, with and without virtualization. The uniform computer environment provided by virtual machines eliminated inconsistent DOCK VS results caused by heterogeneous clusters, however, the execution time for the DOCK VS increased. In our particular experiments, overhead costs were found to be an average of 41% and 2% in execution time for two different clusters, while the actual magnitudes of the execution time costs were minimal. Despite the increase in overhead, virtual clusters are an ideal solution for Grid heterogeneity. With greater development of virtual cluster technology in Grid environments, the problem of platform heterogeneity may be eliminated through virtualization, allowing greater usage of VS, and will benefit all Grid applications in general.

Constrained clusters of gene expression profiles with pathological features.

PubMed

Sese, Jun; Kurokawa, Yukinori; Monden, Morito; Kato, Kikuya; Morishita, Shinichi

2004-11-22

Gene expression profiles should be useful in distinguishing variations in disease, since they reflect accurately the status of cells. The primary clustering of gene expression reveals the genotypes that are responsible for the proximity of members within each cluster, while further clustering elucidates the pathological features of the individual members of each cluster. However, since the first clustering process and the second classification step, in which the features are associated with clusters, are performed independently, the initial set of clusters may omit genes that are associated with pathologically meaningful features. Therefore, it is important to devise a way of identifying gene expression clusters that are associated with pathological features. We present the novel technique of 'itemset constrained clustering' (IC-Clustering), which computes the optimal cluster that maximizes the interclass variance of gene expression between groups, which are divided according to the restriction that only divisions that can be expressed using common features are allowed. This constraint automatically labels each cluster with a set of pathological features which characterize that cluster. When applied to liver cancer datasets, IC-Clustering revealed informative gene expression clusters, which could be annotated with various pathological features, such as 'tumor' and 'man', or 'except tumor' and 'normal liver function'. In contrast, the k-means method overlooked these clusters.

Computational strategies for three-dimensional flow simulations on distributed computer systems. Ph.D. Thesis Semiannual Status Report, 15 Aug. 1993 - 15 Feb. 1994

NASA Technical Reports Server (NTRS)

Weed, Richard Allen; Sankar, L. N.

1994-01-01

An increasing amount of research activity in computational fluid dynamics has been devoted to the development of efficient algorithms for parallel computing systems. The increasing performance to price ratio of engineering workstations has led to research to development procedures for implementing a parallel computing system composed of distributed workstations. This thesis proposal outlines an ongoing research program to develop efficient strategies for performing three-dimensional flow analysis on distributed computing systems. The PVM parallel programming interface was used to modify an existing three-dimensional flow solver, the TEAM code developed by Lockheed for the Air Force, to function as a parallel flow solver on clusters of workstations. Steady flow solutions were generated for three different wing and body geometries to validate the code and evaluate code performance. The proposed research will extend the parallel code development to determine the most efficient strategies for unsteady flow simulations.

Biological and Computational Modeling of Mammographic Density and Stromal Patterning

DTIC Science & Technology

2010-07-01

clumping Score Monolayer Absent Many Absent Absent Absent 1 Nucl. overlap Mild Moderate Mild Micro- nucleoli Rare 2 Clustering Moderate...Few Moderate Micro- nucleoli Occasional 3 Loss cohesion Conspicuous Absent Frequent Macro- nucleoli Frequent 4 We performed serial RPFNA

Facilitating arrhythmia simulation: the method of quantitative cellular automata modeling and parallel running

PubMed Central

Zhu, Hao; Sun, Yan; Rajagopal, Gunaretnam; Mondry, Adrian; Dhar, Pawan

2004-01-01

Background Many arrhythmias are triggered by abnormal electrical activity at the ionic channel and cell level, and then evolve spatio-temporally within the heart. To understand arrhythmias better and to diagnose them more precisely by their ECG waveforms, a whole-heart model is required to explore the association between the massively parallel activities at the channel/cell level and the integrative electrophysiological phenomena at organ level. Methods We have developed a method to build large-scale electrophysiological models by using extended cellular automata, and to run such models on a cluster of shared memory machines. We describe here the method, including the extension of a language-based cellular automaton to implement quantitative computing, the building of a whole-heart model with Visible Human Project data, the parallelization of the model on a cluster of shared memory computers with OpenMP and MPI hybrid programming, and a simulation algorithm that links cellular activity with the ECG. Results We demonstrate that electrical activities at channel, cell, and organ levels can be traced and captured conveniently in our extended cellular automaton system. Examples of some ECG waveforms simulated with a 2-D slice are given to support the ECG simulation algorithm. A performance evaluation of the 3-D model on a four-node cluster is also given. Conclusions Quantitative multicellular modeling with extended cellular automata is a highly efficient and widely applicable method to weave experimental data at different levels into computational models. This process can be used to investigate complex and collective biological activities that can be described neither by their governing differentiation equations nor by discrete parallel computation. Transparent cluster computing is a convenient and effective method to make time-consuming simulation feasible. Arrhythmias, as a typical case, can be effectively simulated with the methods described. PMID:15339335

DID THE INFANT R136 AND NGC 3603 CLUSTERS UNDERGO RESIDUAL GAS EXPULSION?

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

Banerjee, Sambaran; Kroupa, Pavel, E-mail: sambaran@astro.uni-bonn.de, E-mail: pavel@astro.uni-bonn.de

2013-02-10

Based on kinematic data observed for very young, massive clusters that appear to be in dynamical equilibrium, it has recently been argued that such young systems are examples of where the early residual gas expulsion did not happen or had no dynamical effect. The intriguing scenario of a star cluster forming through a single starburst has thereby been challenged. Choosing the case of the R136 cluster of the Large Magellanic Cloud, the most cited one in this context, we perform direct N-body computations that mimic the early evolution of this cluster including the gas-removal phase (on a thermal timescale). Ourmore » calculations show that under plausible initial conditions which are consistent with observational data, a large fraction (>60%) of a gas-expelled, expanding R136-like cluster is bound to regain dynamical equilibrium by its current age. Therefore, the recent measurements of velocity dispersion in the inner regions of R136, which indicate that the cluster is in dynamical equilibrium, are consistent with an earlier substantial gas expulsion of R136 followed by a rapid re-virialization (in Almost-Equal-To 1 Myr). Additionally, we find that the less massive Galactic NGC 3603 Young Cluster (NYC), with a substantially longer re-virialization time, is likely to be found to have deviated from dynamical equilibrium at its present age ( Almost-Equal-To 1 Myr). The recently obtained stellar proper motions in the central part of the NYC indeed suggest this and are consistent with the computed models. This work significantly extends previous models of the Orion Nebula Cluster which already demonstrated that the re-virialization time of young post-gas-expulsion clusters decreases with increasing pre-expulsion density.« less

Did the Infant R136 and NGC 3603 Clusters Undergo Residual Gas Expulsion?

NASA Astrophysics Data System (ADS)

Banerjee, Sambaran; Kroupa, Pavel

2013-02-01

Based on kinematic data observed for very young, massive clusters that appear to be in dynamical equilibrium, it has recently been argued that such young systems are examples of where the early residual gas expulsion did not happen or had no dynamical effect. The intriguing scenario of a star cluster forming through a single starburst has thereby been challenged. Choosing the case of the R136 cluster of the Large Magellanic Cloud, the most cited one in this context, we perform direct N-body computations that mimic the early evolution of this cluster including the gas-removal phase (on a thermal timescale). Our calculations show that under plausible initial conditions which are consistent with observational data, a large fraction (>60%) of a gas-expelled, expanding R136-like cluster is bound to regain dynamical equilibrium by its current age. Therefore, the recent measurements of velocity dispersion in the inner regions of R136, which indicate that the cluster is in dynamical equilibrium, are consistent with an earlier substantial gas expulsion of R136 followed by a rapid re-virialization (in ≈1 Myr). Additionally, we find that the less massive Galactic NGC 3603 Young Cluster (NYC), with a substantially longer re-virialization time, is likely to be found to have deviated from dynamical equilibrium at its present age (≈1 Myr). The recently obtained stellar proper motions in the central part of the NYC indeed suggest this and are consistent with the computed models. This work significantly extends previous models of the Orion Nebula Cluster which already demonstrated that the re-virialization time of young post-gas-expulsion clusters decreases with increasing pre-expulsion density.

Self-consistent Green's function embedding for advanced electronic structure methods based on a dynamical mean-field concept

NASA Astrophysics Data System (ADS)

Chibani, Wael; Ren, Xinguo; Scheffler, Matthias; Rinke, Patrick

2016-04-01

We present an embedding scheme for periodic systems that facilitates the treatment of the physically important part (here a unit cell or a supercell) with advanced electronic structure methods, that are computationally too expensive for periodic systems. The rest of the periodic system is treated with computationally less demanding approaches, e.g., Kohn-Sham density-functional theory, in a self-consistent manner. Our scheme is based on the concept of dynamical mean-field theory formulated in terms of Green's functions. Our real-space dynamical mean-field embedding scheme features two nested Dyson equations, one for the embedded cluster and another for the periodic surrounding. The total energy is computed from the resulting Green's functions. The performance of our scheme is demonstrated by treating the embedded region with hybrid functionals and many-body perturbation theory in the GW approach for simple bulk systems. The total energy and the density of states converge rapidly with respect to the computational parameters and approach their bulk limit with increasing cluster (i.e., computational supercell) size.

Templet Web: the use of volunteer computing approach in PaaS-style cloud

NASA Astrophysics Data System (ADS)

Vostokin, Sergei; Artamonov, Yuriy; Tsarev, Daniil

2018-03-01

This article presents the Templet Web cloud service. The service is designed for high-performance scientific computing automation. The use of high-performance technology is specifically required by new fields of computational science such as data mining, artificial intelligence, machine learning, and others. Cloud technologies provide a significant cost reduction for high-performance scientific applications. The main objectives to achieve this cost reduction in the Templet Web service design are: (a) the implementation of "on-demand" access; (b) source code deployment management; (c) high-performance computing programs development automation. The distinctive feature of the service is the approach mainly used in the field of volunteer computing, when a person who has access to a computer system delegates his access rights to the requesting user. We developed an access procedure, algorithms, and software for utilization of free computational resources of the academic cluster system in line with the methods of volunteer computing. The Templet Web service has been in operation for five years. It has been successfully used for conducting laboratory workshops and solving research problems, some of which are considered in this article. The article also provides an overview of research directions related to service development.

Parallel programming of gradient-based iterative image reconstruction schemes for optical tomography.

PubMed

Hielscher, Andreas H; Bartel, Sebastian

2004-02-01

Optical tomography (OT) is a fast developing novel imaging modality that uses near-infrared (NIR) light to obtain cross-sectional views of optical properties inside the human body. A major challenge remains the time-consuming, computational-intensive image reconstruction problem that converts NIR transmission measurements into cross-sectional images. To increase the speed of iterative image reconstruction schemes that are commonly applied for OT, we have developed and implemented several parallel algorithms on a cluster of workstations. Static process distribution as well as dynamic load balancing schemes suitable for heterogeneous clusters and varying machine performances are introduced and tested. The resulting algorithms are shown to accelerate the reconstruction process to various degrees, substantially reducing the computation times for clinically relevant problems.

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.

A First Estimate of the X-Ray Binary Frequency as a Function of Star Cluster Mass in a Single Galactic System

NASA Astrophysics Data System (ADS)

Clark, D. M.; Eikenberry, S. S.; Brandl, B. R.; Wilson, J. C.; Carson, J. C.; Henderson, C. P.; Hayward, T. L.; Barry, D. J.; Ptak, A. F.; Colbert, E. J. M.

2008-05-01

We use the previously identified 15 infrared star cluster counterparts to X-ray point sources in the interacting galaxies NGC 4038/4039 (the Antennae) to study the relationship between total cluster mass and X-ray binary number. This significant population of X-Ray/IR associations allows us to perform, for the first time, a statistical study of X-ray point sources and their environments. We define a quantity, η, relating the fraction of X-ray sources per unit mass as a function of cluster mass in the Antennae. We compute cluster mass by fitting spectral evolutionary models to Ks luminosity. Considering that this method depends on cluster age, we use four different age distributions to explore the effects of cluster age on the value of η and find it varies by less than a factor of 4. We find a mean value of η for these different distributions of η = 1.7 × 10-8 M-1⊙ with ση = 1.2 × 10-8 M-1⊙. Performing a χ2 test, we demonstrate η could exhibit a positive slope, but that it depends on the assumed distribution in cluster ages. While the estimated uncertainties in η are factors of a few, we believe this is the first estimate made of this quantity to "order of magnitude" accuracy. We also compare our findings to theoretical models of open and globular cluster evolution, incorporating the X-ray binary fraction per cluster.

Agent-based method for distributed clustering of textual information

DOEpatents

Potok, Thomas E [Oak Ridge, TN; Reed, Joel W [Knoxville, TN; Elmore, Mark T [Oak Ridge, TN; Treadwell, Jim N [Louisville, TN

2010-09-28

A computer method and system for storing, retrieving and displaying information has a multiplexing agent (20) that calculates a new document vector (25) for a new document (21) to be added to the system and transmits the new document vector (25) to master cluster agents (22) and cluster agents (23) for evaluation. These agents (22, 23) perform the evaluation and return values upstream to the multiplexing agent (20) based on the similarity of the document to documents stored under their control. The multiplexing agent (20) then sends the document (21) and the document vector (25) to the master cluster agent (22), which then forwards it to a cluster agent (23) or creates a new cluster agent (23) to manage the document (21). The system also searches for stored documents according to a search query having at least one term and identifying the documents found in the search, and displays the documents in a clustering display (80) of similarity so as to indicate similarity of the documents to each other.

FOSS GIS on the GFZ HPC cluster: Towards a service-oriented Scientific Geocomputation Environment

NASA Astrophysics Data System (ADS)

Loewe, P.; Klump, J.; Thaler, J.

2012-12-01

High performance compute clusters can be used as geocomputation workbenches. Their wealth of resources enables us to take on geocomputation tasks which exceed the limitations of smaller systems. These general capabilities can be harnessed via tools such as Geographic Information System (GIS), provided they are able to utilize the available cluster configuration/architecture and provide a sufficient degree of user friendliness to allow for wide application. While server-level computing is clearly not sufficient for the growing numbers of data- or computation-intense tasks undertaken, these tasks do not get even close to the requirements needed for access to "top shelf" national cluster facilities. So until recently such kind of geocomputation research was effectively barred due to lack access to of adequate resources. In this paper we report on the experiences gained by providing GRASS GIS as a software service on a HPC compute cluster at the German Research Centre for Geosciences using Platform Computing's Load Sharing Facility (LSF). GRASS GIS is the oldest and largest Free Open Source (FOSS) GIS project. During ramp up in 2011, multiple versions of GRASS GIS (v 6.4.2, 6.5 and 7.0) were installed on the HPC compute cluster, which currently consists of 234 nodes with 480 CPUs providing 3084 cores. Nineteen different processing queues with varying hardware capabilities and priorities are provided, allowing for fine-grained scheduling and load balancing. After successful initial testing, mechanisms were developed to deploy scripted geocomputation tasks onto dedicated processing queues. The mechanisms are based on earlier work by NETELER et al. (2008) and allow to use all 3084 cores for GRASS based geocomputation work. However, in practice applications are limited to fewer resources as assigned to their respective queue. Applications of the new GIS functionality comprise so far of hydrological analysis, remote sensing and the generation of maps of simulated tsunamis in the Mediterranean Sea for the Tsunami Atlas of the FP-7 TRIDEC Project (www.tridec-online.eu). This included the processing of complex problems, requiring significant amounts of processing time up to full 20 CPU days. This GRASS GIS-based service is provided as a research utility in the sense of "Software as a Service" (SaaS) and is a first step towards a GFZ corporate cloud service.

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.

HPC on Competitive Cloud Resources

NASA Astrophysics Data System (ADS)

Bientinesi, Paolo; Iakymchuk, Roman; Napper, Jeff

Computing as a utility has reached the mainstream. Scientists can now easily rent time on large commercial clusters that can be expanded and reduced on-demand in real-time. However, current commercial cloud computing performance falls short of systems specifically designed for scientific applications. Scientific computing needs are quite different from those of the web applications that have been the focus of cloud computing vendors. In this chapter we demonstrate through empirical evaluation the computational efficiency of high-performance numerical applications in a commercial cloud environment when resources are shared under high contention. Using the Linpack benchmark as a case study, we show that cache utilization becomes highly unpredictable and similarly affects computation time. For some problems, not only is it more efficient to underutilize resources, but the solution can be reached sooner in realtime (wall-time). We also show that the smallest, cheapest (64-bit) instance on the studied environment is the best for price to performance ration. In light of the high-contention we witness, we believe that alternative definitions of efficiency for commercial cloud environments should be introduced where strong performance guarantees do not exist. Concepts like average, expected performance and execution time, expected cost to completion, and variance measures--traditionally ignored in the high-performance computing context--now should complement or even substitute the standard definitions of efficiency.

An Innovative Approach to Bridge a Skill Gap and Grow a Workforce Pipeline: The Computer System, Cluster, and Networking Summer Institute

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

Connor, Carolyn Marie; Jacobson, Andree Lars; Bonnie, Amanda Marie

Sustainable and effective computing infrastructure depends critically on the skills and expertise of domain scientists and of committed and well-trained advanced computing professionals. But, in its ongoing High Performance Computing (HPC) work, Los Alamos National Laboratory noted a persistent shortage of well-prepared applicants, particularly for entry-level cluster administration, file systems administration, and high speed networking positions. Further, based upon recruiting efforts and interactions with universities graduating students in related majors of interest (e.g., computer science (CS)), there has been a long standing skillset gap, as focused training in HPC topics is typically lacking or absent in undergraduate and in evenmore » many graduate programs. Given that the effective operation and use of HPC systems requires specialized and often advanced training, that there is a recognized HPC skillset gap, and that there is intense global competition for computing and computational science talent, there is a long-standing and critical need for innovative approaches to help bridge the gap and create a well-prepared, next generation HPC workforce. Our paper places this need in the context of the HPC work and workforce requirements at Los Alamos National Laboratory (LANL) and presents one such innovative program conceived to address the need, bridge the gap, and grow an HPC workforce pipeline at LANL. The Computer System, Cluster, and Networking Summer Institute (CSCNSI) completed its 10th year in 2016. The story of the CSCNSI and its evolution is detailed below with a description of the design of its Boot Camp, and a summary of its success and some key factors that have enabled that success.« less

An Innovative Approach to Bridge a Skill Gap and Grow a Workforce Pipeline: The Computer System, Cluster, and Networking Summer Institute

DOE PAGES

Connor, Carolyn Marie; Jacobson, Andree Lars; Bonnie, Amanda Marie; ...

2016-11-01

Sustainable and effective computing infrastructure depends critically on the skills and expertise of domain scientists and of committed and well-trained advanced computing professionals. But, in its ongoing High Performance Computing (HPC) work, Los Alamos National Laboratory noted a persistent shortage of well-prepared applicants, particularly for entry-level cluster administration, file systems administration, and high speed networking positions. Further, based upon recruiting efforts and interactions with universities graduating students in related majors of interest (e.g., computer science (CS)), there has been a long standing skillset gap, as focused training in HPC topics is typically lacking or absent in undergraduate and in evenmore » many graduate programs. Given that the effective operation and use of HPC systems requires specialized and often advanced training, that there is a recognized HPC skillset gap, and that there is intense global competition for computing and computational science talent, there is a long-standing and critical need for innovative approaches to help bridge the gap and create a well-prepared, next generation HPC workforce. Our paper places this need in the context of the HPC work and workforce requirements at Los Alamos National Laboratory (LANL) and presents one such innovative program conceived to address the need, bridge the gap, and grow an HPC workforce pipeline at LANL. The Computer System, Cluster, and Networking Summer Institute (CSCNSI) completed its 10th year in 2016. The story of the CSCNSI and its evolution is detailed below with a description of the design of its Boot Camp, and a summary of its success and some key factors that have enabled that success.« less

A non-voxel-based broad-beam (NVBB) framework for IMRT treatment planning.

PubMed

Lu, Weiguo

2010-12-07

We present a novel framework that enables very large scale intensity-modulated radiation therapy (IMRT) planning in limited computation resources with improvements in cost, plan quality and planning throughput. Current IMRT optimization uses a voxel-based beamlet superposition (VBS) framework that requires pre-calculation and storage of a large amount of beamlet data, resulting in large temporal and spatial complexity. We developed a non-voxel-based broad-beam (NVBB) framework for IMRT capable of direct treatment parameter optimization (DTPO). In this framework, both objective function and derivative are evaluated based on the continuous viewpoint, abandoning 'voxel' and 'beamlet' representations. Thus pre-calculation and storage of beamlets are no longer needed. The NVBB framework has linear complexities (O(N(3))) in both space and time. The low memory, full computation and data parallelization nature of the framework render its efficient implementation on the graphic processing unit (GPU). We implemented the NVBB framework and incorporated it with the TomoTherapy treatment planning system (TPS). The new TPS runs on a single workstation with one GPU card (NVBB-GPU). Extensive verification/validation tests were performed in house and via third parties. Benchmarks on dose accuracy, plan quality and throughput were compared with the commercial TomoTherapy TPS that is based on the VBS framework and uses a computer cluster with 14 nodes (VBS-cluster). For all tests, the dose accuracy of these two TPSs is comparable (within 1%). Plan qualities were comparable with no clinically significant difference for most cases except that superior target uniformity was seen in the NVBB-GPU for some cases. However, the planning time using the NVBB-GPU was reduced many folds over the VBS-cluster. In conclusion, we developed a novel NVBB framework for IMRT optimization. The continuous viewpoint and DTPO nature of the algorithm eliminate the need for beamlets and lead to better plan quality. The computation parallelization on a GPU instead of a computer cluster significantly reduces hardware and service costs. Compared with using the current VBS framework on a computer cluster, the planning time is significantly reduced using the NVBB framework on a single workstation with a GPU card.

Inductive Approaches to Improving Diagnosis and Design for Diagnosability

NASA Technical Reports Server (NTRS)

Fisher, Douglas H. (Principal Investigator)

1995-01-01

The first research area under this grant addresses the problem of classifying time series according to their morphological features in the time domain. A supervised learning system called CALCHAS, which induces a classification procedure for signatures from preclassified examples, was developed. For each of several signature classes, the system infers a model that captures the class's morphological features using Bayesian model induction and the minimum message length approach to assign priors. After induction, a time series (signature) is classified in one of the classes when there is enough evidence to support that decision. Time series with sufficiently novel features, belonging to classes not present in the training set, are recognized as such. A second area of research assumes two sources of information about a system: a model or domain theory that encodes aspects of the system under study and data from actual system operations over time. A model, when it exists, represents strong prior expectations about how a system will perform. Our work with a diagnostic model of the RCS (Reaction Control System) of the Space Shuttle motivated the development of SIG, a system which combines information from a model (or domain theory) and data. As it tracks RCS behavior, the model computes quantitative and qualitative values. Induction is then performed over the data represented by both the 'raw' features and the model-computed high-level features. Finally, work on clustering for operating mode discovery motivated some important extensions to the clustering strategy we had used. One modification appends an iterative optimization technique onto the clustering system; this optimization strategy appears to be novel in the clustering literature. A second modification improves the noise tolerance of the clustering system. In particular, we adapt resampling-based pruning strategies used by supervised learning systems to the task of simplifying hierarchical clusterings, thus making post-clustering analysis easier.

The Gap Procedure: for the identification of phylogenetic clusters in HIV-1 sequence data.

PubMed

Vrbik, Irene; Stephens, David A; Roger, Michel; Brenner, Bluma G

2015-11-04

In the context of infectious disease, sequence clustering can be used to provide important insights into the dynamics of transmission. Cluster analysis is usually performed using a phylogenetic approach whereby clusters are assigned on the basis of sufficiently small genetic distances and high bootstrap support (or posterior probabilities). The computational burden involved in this phylogenetic threshold approach is a major drawback, especially when a large number of sequences are being considered. In addition, this method requires a skilled user to specify the appropriate threshold values which may vary widely depending on the application. This paper presents the Gap Procedure, a distance-based clustering algorithm for the classification of DNA sequences sampled from individuals infected with the human immunodeficiency virus type 1 (HIV-1). Our heuristic algorithm bypasses the need for phylogenetic reconstruction, thereby supporting the quick analysis of large genetic data sets. Moreover, this fully automated procedure relies on data-driven gaps in sorted pairwise distances to infer clusters, thus no user-specified threshold values are required. The clustering results obtained by the Gap Procedure on both real and simulated data, closely agree with those found using the threshold approach, while only requiring a fraction of the time to complete the analysis. Apart from the dramatic gains in computational time, the Gap Procedure is highly effective in finding distinct groups of genetically similar sequences and obviates the need for subjective user-specified values. The clusters of genetically similar sequences returned by this procedure can be used to detect patterns in HIV-1 transmission and thereby aid in the prevention, treatment and containment of the disease.

Fault-tolerant quantum computation with nondeterministic entangling gates

NASA Astrophysics Data System (ADS)

Auger, James M.; Anwar, Hussain; Gimeno-Segovia, Mercedes; Stace, Thomas M.; Browne, Dan E.

2018-03-01

Performing entangling gates between physical qubits is necessary for building a large-scale universal quantum computer, but in some physical implementations—for example, those that are based on linear optics or networks of ion traps—entangling gates can only be implemented probabilistically. In this work, we study the fault-tolerant performance of a topological cluster state scheme with local nondeterministic entanglement generation, where failed entangling gates (which correspond to bonds on the lattice representation of the cluster state) lead to a defective three-dimensional lattice with missing bonds. We present two approaches for dealing with missing bonds; the first is a nonadaptive scheme that requires no additional quantum processing, and the second is an adaptive scheme in which qubits can be measured in an alternative basis to effectively remove them from the lattice, hence eliminating their damaging effect and leading to better threshold performance. We find that a fault-tolerance threshold can still be observed with a bond-loss rate of 6.5% for the nonadaptive scheme, and a bond-loss rate as high as 14.5% for the adaptive scheme.

Implementing Journaling in a Linux Shared Disk File System

NASA Technical Reports Server (NTRS)

Preslan, Kenneth W.; Barry, Andrew; Brassow, Jonathan; Cattelan, Russell; Manthei, Adam; Nygaard, Erling; VanOort, Seth; Teigland, David; Tilstra, Mike; O'Keefe, Matthew;

Accelerating Subsurface Transport Simulation on Heterogeneous Clusters

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

Villa, Oreste; Gawande, Nitin A.; Tumeo, Antonino

Reactive transport numerical models simulate chemical and microbiological reactions that occur along a flowpath. These models have to compute reactions for a large number of locations. They solve the set of ordinary differential equations (ODEs) that describes the reaction for each location through the Newton-Raphson technique. This technique involves computing a Jacobian matrix and a residual vector for each set of equation, and then solving iteratively the linearized system by performing Gaussian Elimination and LU decomposition until convergence. STOMP, a well known subsurface flow simulation tool, employs matrices with sizes in the order of 100x100 elements and, for numerical accuracy,more » LU factorization with full pivoting instead of the faster partial pivoting. Modern high performance computing systems are heterogeneous machines whose nodes integrate both CPUs and GPUs, exposing unprecedented amounts of parallelism. To exploit all their computational power, applications must use both the types of processing elements. For the case of subsurface flow simulation, this mainly requires implementing efficient batched LU-based solvers and identifying efficient solutions for enabling load balancing among the different processors of the system. In this paper we discuss two approaches that allows scaling STOMP's performance on heterogeneous clusters. We initially identify the challenges in implementing batched LU-based solvers for small matrices on GPUs, and propose an implementation that fulfills STOMP's requirements. We compare this implementation to other existing solutions. Then, we combine the batched GPU solver with an OpenMP-based CPU solver, and present an adaptive load balancer that dynamically distributes the linear systems to solve between the two components inside a node. We show how these approaches, integrated into the full application, provide speed ups from 6 to 7 times on large problems, executed on up to 16 nodes of a cluster with two AMD Opteron 6272 and a Tesla M2090 per node.« less

Clustering of dietary intake and sedentary behavior in 2-year-old children.

PubMed

Gubbels, Jessica S; Kremers, Stef P J; Stafleu, Annette; Dagnelie, Pieter C; de Vries, Sanne I; de Vries, Nanne K; Thijs, Carel

2009-08-01

To examine clustering of energy balance-related behaviors (EBRBs) in young children. This is crucial because lifestyle habits are formed at an early age and track in later life. This study is the first to examine EBRB clustering in children as young as 2 years. Cross-sectional data originated from the Child, Parent and Health: Lifestyle and Genetic Constitution (KOALA) Birth Cohort Study. Parents of 2578 2-year-old children completed a questionnaire. Correlation analyses, principal component analyses, and linear regression analyses were performed to examine clustering of EBRBs. We found modest but consistent correlations in EBRBs. Two clusters emerged: a "sedentary-snacking cluster" and a "fiber cluster." Television viewing clustered with computer use and unhealthy dietary behaviors. Children who frequently consumed vegetables also consumed fruit and brown bread more often and white bread less often. Lower maternal education and maternal obesity were associated with high scores on the sedentary-snacking cluster, whereas higher educational level was associated with high fiber cluster scores. Obesity-prone behavioral clusters are already visible in 2-year-old children and are related to maternal characteristics. The findings suggest that obesity prevention should apply an integrated approach to physical activity and dietary intake in early childhood.

Promoting Interests in Atmospheric Science at a Liberal Arts Institution

NASA Astrophysics Data System (ADS)

Roussev, S.; Sherengos, P. M.; Limpasuvan, V.; Xue, M.

2007-12-01

Coastal Carolina University (CCU) students in Computer Science participated in a project to set up an operational weather forecast for the local community. The project involved the construction of two computing clusters and the automation of daily forecasting. Funded by NSF-MRI, two high-performance clusters were successfully established to run the University of Oklahoma's Advance Regional Prediction System (ARPS). Daily weather predictions are made over South Carolina and North Carolina at 3-km horizontal resolution (roughly 1.9 miles) using initial and boundary condition data provided by UNIDATA. At this high resolution, the model is cloud- resolving, thus providing detailed picture of heavy thunderstorms and precipitation. Forecast results are displayed on CCU's website (https://marc.coastal.edu/HPC) to complement observations at the National Weather Service in Wilmington N.C. Present efforts include providing forecasts at 1-km resolution (or finer), comparisons with other models like Weather Research and Forecasting (WRF) model, and the examination of local phenomena (like water spouts and tornadoes). Through these activities the students learn about shell scripting, cluster operating systems, and web design. More importantly, students are introduced to Atmospheric Science, the processes involved in making weather forecasts, and the interpretation of their forecasts. Simulations generated by the forecasts will be integrated into the contents of CCU's course like Fluid Dynamics, Atmospheric Sciences, Atmospheric Physics, and Remote Sensing. Operated jointly between the departments of Applied Physics and Computer Science, the clusters are expected to be used by CCU faculty and students for future research and inquiry-based projects in Computer Science, Applied Physics, and Marine Science.

SLUG - stochastically lighting up galaxies - III. A suite of tools for simulated photometry, spectroscopy, and Bayesian inference with stochastic stellar populations

NASA Astrophysics Data System (ADS)

Krumholz, Mark R.; Fumagalli, Michele; da Silva, Robert L.; Rendahl, Theodore; Parra, Jonathan

2015-09-01

Stellar population synthesis techniques for predicting the observable light emitted by a stellar population have extensive applications in numerous areas of astronomy. However, accurate predictions for small populations of young stars, such as those found in individual star clusters, star-forming dwarf galaxies, and small segments of spiral galaxies, require that the population be treated stochastically. Conversely, accurate deductions of the properties of such objects also require consideration of stochasticity. Here we describe a comprehensive suite of modular, open-source software tools for tackling these related problems. These include the following: a greatly-enhanced version of the SLUG code introduced by da Silva et al., which computes spectra and photometry for stochastically or deterministically sampled stellar populations with nearly arbitrary star formation histories, clustering properties, and initial mass functions; CLOUDY_SLUG, a tool that automatically couples SLUG-computed spectra with the CLOUDY radiative transfer code in order to predict stochastic nebular emission; BAYESPHOT, a general-purpose tool for performing Bayesian inference on the physical properties of stellar systems based on unresolved photometry; and CLUSTER_SLUG and SFR_SLUG, a pair of tools that use BAYESPHOT on a library of SLUG models to compute the mass, age, and extinction of mono-age star clusters, and the star formation rate of galaxies, respectively. The latter two tools make use of an extensive library of pre-computed stellar population models, which are included in the software. The complete package is available at http://www.slugsps.com.

Mapping of terrain by computer clustering techniques using multispectral scanner data and using color aerial film

NASA Technical Reports Server (NTRS)

Smedes, H. W.; Linnerud, H. J.; Woolaver, L. B.; Su, M. Y.; Jayroe, R. R.

1972-01-01

Two clustering techniques were used for terrain mapping by computer of test sites in Yellowstone National Park. One test was made with multispectral scanner data using a composite technique which consists of (1) a strictly sequential statistical clustering which is a sequential variance analysis, and (2) a generalized K-means clustering. In this composite technique, the output of (1) is a first approximation of the cluster centers. This is the input to (2) which consists of steps to improve the determination of cluster centers by iterative procedures. Another test was made using the three emulsion layers of color-infrared aerial film as a three-band spectrometer. Relative film densities were analyzed using a simple clustering technique in three-color space. Important advantages of the clustering technique over conventional supervised computer programs are (1) human intervention, preparation time, and manipulation of data are reduced, (2) the computer map, gives unbiased indication of where best to select the reference ground control data, (3) use of easy to obtain inexpensive film, and (4) the geometric distortions can be easily rectified by simple standard photogrammetric techniques.

SANs and Large Scale Data Migration at the NASA Center for Computational Sciences

NASA Technical Reports Server (NTRS)

Salmon, Ellen M.

2004-01-01

Evolution and migration are a way of life for provisioners of high-performance mass storage systems that serve high-end computers used by climate and Earth and space science researchers: the compute engines come and go, but the data remains. At the NASA Center for Computational Sciences (NCCS), disk and tape SANs are deployed to provide high-speed I/O for the compute engines and the hierarchical storage management systems. Along with gigabit Ethernet, they also enable the NCCS's latest significant migration: the transparent transfer of 300 Til3 of legacy HSM data into the new Sun SAM-QFS cluster.

Precision of systematic and random sampling in clustered populations: habitat patches and aggregating organisms.

PubMed

McGarvey, Richard; Burch, Paul; Matthews, Janet M

2016-01-01

Natural populations of plants and animals spatially cluster because (1) suitable habitat is patchy, and (2) within suitable habitat, individuals aggregate further into clusters of higher density. We compare the precision of random and systematic field sampling survey designs under these two processes of species clustering. Second, we evaluate the performance of 13 estimators for the variance of the sample mean from a systematic survey. Replicated simulated surveys, as counts from 100 transects, allocated either randomly or systematically within the study region, were used to estimate population density in six spatial point populations including habitat patches and Matérn circular clustered aggregations of organisms, together and in combination. The standard one-start aligned systematic survey design, a uniform 10 x 10 grid of transects, was much more precise. Variances of the 10 000 replicated systematic survey mean densities were one-third to one-fifth of those from randomly allocated transects, implying transect sample sizes giving equivalent precision by random survey would need to be three to five times larger. Organisms being restricted to patches of habitat was alone sufficient to yield this precision advantage for the systematic design. But this improved precision for systematic sampling in clustered populations is underestimated by standard variance estimators used to compute confidence intervals. True variance for the survey sample mean was computed from the variance of 10 000 simulated survey mean estimates. Testing 10 published and three newly proposed variance estimators, the two variance estimators (v) that corrected for inter-transect correlation (ν₈ and ν(W)) were the most accurate and also the most precise in clustered populations. These greatly outperformed the two "post-stratification" variance estimators (ν₂ and ν₃) that are now more commonly applied in systematic surveys. Similar variance estimator performance rankings were found with a second differently generated set of spatial point populations, ν₈ and ν(W) again being the best performers in the longer-range autocorrelated populations. However, no systematic variance estimators tested were free from bias. On balance, systematic designs bring more narrow confidence intervals in clustered populations, while random designs permit unbiased estimates of (often wider) confidence interval. The search continues for better estimators of sampling variance for the systematic survey mean.

Copy-number analysis and inference of subclonal populations in cancer genomes using Sclust.

PubMed

Cun, Yupeng; Yang, Tsun-Po; Achter, Viktor; Lang, Ulrich; Peifer, Martin

2018-06-01

The genomes of cancer cells constantly change during pathogenesis. This evolutionary process can lead to the emergence of drug-resistant mutations in subclonal populations, which can hinder therapeutic intervention in patients. Data derived from massively parallel sequencing can be used to infer these subclonal populations using tumor-specific point mutations. The accurate determination of copy-number changes and tumor impurity is necessary to reliably infer subclonal populations by mutational clustering. This protocol describes how to use Sclust, a copy-number analysis method with a recently developed mutational clustering approach. In a series of simulations and comparisons with alternative methods, we have previously shown that Sclust accurately determines copy-number states and subclonal populations. Performance tests show that the method is computationally efficient, with copy-number analysis and mutational clustering taking <10 min. Sclust is designed such that even non-experts in computational biology or bioinformatics with basic knowledge of the Linux/Unix command-line syntax should be able to carry out analyses of subclonal populations.

A new scheme for perturbative triples correction to (0,1) sector of Fock space multi-reference coupled cluster method: theory, implementation, and examples.

PubMed

Dutta, Achintya Kumar; Vaval, Nayana; Pal, Sourav

2015-01-28

We propose a new elegant strategy to implement third order triples correction in the light of many-body perturbation theory to the Fock space multi-reference coupled cluster method for the ionization problem. The computational scaling as well as the storage requirement is of key concerns in any many-body calculations. Our proposed approach scales as N(6) does not require the storage of triples amplitudes and gives superior agreement over all the previous attempts made. This approach is capable of calculating multiple roots in a single calculation in contrast to the inclusion of perturbative triples in the equation of motion variant of the coupled cluster theory, where each root needs to be computed in a state-specific way and requires both the left and right state vectors together. The performance of the newly implemented scheme is tested by applying to methylene, boron nitride (B2N) anion, nitrogen, water, carbon monoxide, acetylene, formaldehyde, and thymine monomer, a DNA base.

MPIGeneNet: Parallel Calculation of Gene Co-Expression Networks on Multicore Clusters.

PubMed

Gonzalez-Dominguez, Jorge; Martin, Maria J

2017-10-10

In this work we present MPIGeneNet, a parallel tool that applies Pearson's correlation and Random Matrix Theory to construct gene co-expression networks. It is based on the state-of-the-art sequential tool RMTGeneNet, which provides networks with high robustness and sensitivity at the expenses of relatively long runtimes for large scale input datasets. MPIGeneNet returns the same results as RMTGeneNet but improves the memory management, reduces the I/O cost, and accelerates the two most computationally demanding steps of co-expression network construction by exploiting the compute capabilities of common multicore CPU clusters. Our performance evaluation on two different systems using three typical input datasets shows that MPIGeneNet is significantly faster than RMTGeneNet. As an example, our tool is up to 175.41 times faster on a cluster with eight nodes, each one containing two 12-core Intel Haswell processors. Source code of MPIGeneNet, as well as a reference manual, are available at https://sourceforge.net/projects/mpigenenet/.

Analysis of the bond-valence method for calculating (29) Si and (31) P magnetic shielding in covalent network solids.

PubMed

Holmes, Sean T; Alkan, Fahri; Iuliucci, Robbie J; Mueller, Karl T; Dybowski, Cecil

2016-07-05

(29) Si and (31) P magnetic-shielding tensors in covalent network solids have been evaluated using periodic and cluster-based calculations. The cluster-based computational methodology employs pseudoatoms to reduce the net charge (resulting from missing co-ordination on the terminal atoms) through valence modification of terminal atoms using bond-valence theory (VMTA/BV). The magnetic-shielding tensors computed with the VMTA/BV method are compared to magnetic-shielding tensors determined with the periodic GIPAW approach. The cluster-based all-electron calculations agree with experiment better than the GIPAW calculations, particularly for predicting absolute magnetic shielding and for predicting chemical shifts. The performance of the DFT functionals CA-PZ, PW91, PBE, rPBE, PBEsol, WC, and PBE0 are assessed for the prediction of (29) Si and (31) P magnetic-shielding constants. Calculations using the hybrid functional PBE0, in combination with the VMTA/BV approach, result in excellent agreement with experiment. © 2016 Wiley Periodicals, Inc. © 2016 Wiley Periodicals, Inc.

A secure protocol for protecting the identity of providers when disclosing data for disease surveillance

PubMed Central

Hu, Jun; Mercer, Jay; Peyton, Liam; Kantarcioglu, Murat; Malin, Bradley; Buckeridge, David; Samet, Saeed; Earle, Craig

2011-01-01

Background Providers have been reluctant to disclose patient data for public-health purposes. Even if patient privacy is ensured, the desire to protect provider confidentiality has been an important driver of this reluctance. Methods Six requirements for a surveillance protocol were defined that satisfy the confidentiality needs of providers and ensure utility to public health. The authors developed a secure multi-party computation protocol using the Paillier cryptosystem to allow the disclosure of stratified case counts and denominators to meet these requirements. The authors evaluated the protocol in a simulated environment on its computation performance and ability to detect disease outbreak clusters. Results Theoretical and empirical assessments demonstrate that all requirements are met by the protocol. A system implementing the protocol scales linearly in terms of computation time as the number of providers is increased. The absolute time to perform the computations was 12.5 s for data from 3000 practices. This is acceptable performance, given that the reporting would normally be done at 24 h intervals. The accuracy of detection disease outbreak cluster was unchanged compared with a non-secure distributed surveillance protocol, with an F-score higher than 0.92 for outbreaks involving 500 or more cases. Conclusion The protocol and associated software provide a practical method for providers to disclose patient data for sentinel, syndromic or other indicator-based surveillance while protecting patient privacy and the identity of individual providers. PMID:21486880

Towards the use of computationally inserted lesions for mammographic CAD assessment

NASA Astrophysics Data System (ADS)

Ghanian, Zahra; Pezeshk, Aria; Petrick, Nicholas; Sahiner, Berkman

2018-03-01

Computer-aided detection (CADe) devices used for breast cancer detection on mammograms are typically first developed and assessed for a specific "original" acquisition system, e.g., a specific image detector. When CADe developers are ready to apply their CADe device to a new mammographic acquisition system, they typically assess the CADe device with images acquired using the new system. Collecting large repositories of clinical images containing verified cancer locations and acquired by the new image acquisition system is costly and time consuming. Our goal is to develop a methodology to reduce the clinical data burden in the assessment of a CADe device for use with a different image acquisition system. We are developing an image blending technique that allows users to seamlessly insert lesions imaged using an original acquisition system into normal images or regions acquired with a new system. In this study, we investigated the insertion of microcalcification clusters imaged using an original acquisition system into normal images acquired with that same system utilizing our previously-developed image blending technique. We first performed a reader study to assess whether experienced observers could distinguish between computationally inserted and native clusters. For this purpose, we applied our insertion technique to clinical cases taken from the University of South Florida Digital Database for Screening Mammography (DDSM) and the Breast Cancer Digital Repository (BCDR). Regions of interest containing microcalcification clusters from one breast of a patient were inserted into the contralateral breast of the same patient. The reader study included 55 native clusters and their 55 inserted counterparts. Analysis of the reader ratings using receiver operating characteristic (ROC) methodology indicated that inserted clusters cannot be reliably distinguished from native clusters (area under the ROC curve, AUC=0.58±0.04). Furthermore, CADe sensitivity was evaluated on mammograms with native and inserted microcalcification clusters using a commercial CADe system. For this purpose, we used full field digital mammograms (FFDMs) from 68 clinical cases, acquired at the University of Michigan Health System. The average sensitivities for native and inserted clusters were equal, 85.3% (58/68). These results demonstrate the feasibility of using the inserted microcalcification clusters for assessing mammographic CAD devices.

Computing the cross sections of nuclear reactions with nuclear clusters emission for proton energies between 30 MeV and 2.6 GeV

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

Korovin, Yu. A.; Maksimushkina, A. V., E-mail: AVMaksimushkina@mephi.ru; Frolova, T. A.

2016-12-15

The cross sections of nuclear reactions involving emission of clusters of light nuclei in proton collisions with a heavy-metal target are computed for incident-proton energies between 30 MeV and 2.6 GeV. The calculation relies on the ALICE/ASH and CASCADE/INPE computer codes. The parameters determining the pre-equilibrium cluster emission are varied in the computation.

Algorithms and software used in selecting structure of machine-training cluster based on neurocomputers

NASA Astrophysics Data System (ADS)

Romanchuk, V. A.; Lukashenko, V. V.

2018-05-01

The technique of functioning of a control system by a computing cluster based on neurocomputers is proposed. Particular attention is paid to the method of choosing the structure of the computing cluster due to the fact that the existing methods are not effective because of a specialized hardware base - neurocomputers, which are highly parallel computer devices with an architecture different from the von Neumann architecture. A developed algorithm for choosing the computational structure of a cloud cluster is described, starting from the direction of data transfer in the flow control graph of the program and its adjacency matrix.

Scalable Prediction of Energy Consumption using Incremental Time Series Clustering

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

Simmhan, Yogesh; Noor, Muhammad Usman

2013-10-09

Time series datasets are a canonical form of high velocity Big Data, and often generated by pervasive sensors, such as found in smart infrastructure. Performing predictive analytics on time series data can be computationally complex, and requires approximation techniques. In this paper, we motivate this problem using a real application from the smart grid domain. We propose an incremental clustering technique, along with a novel affinity score for determining cluster similarity, which help reduce the prediction error for cumulative time series within a cluster. We evaluate this technique, along with optimizations, using real datasets from smart meters, totaling ~700,000 datamore » points, and show the efficacy of our techniques in improving the prediction error of time series data within polynomial time.« less

A theoretical and experimental benchmark study of core-excited states in nitrogen

NASA Astrophysics Data System (ADS)

Myhre, Rolf H.; Wolf, Thomas J. A.; Cheng, Lan; Nandi, Saikat; Coriani, Sonia; Gühr, Markus; Koch, Henrik

2018-02-01

The high resolution near edge X-ray absorption fine structure spectrum of nitrogen displays the vibrational structure of the core-excited states. This makes nitrogen well suited for assessing the accuracy of different electronic structure methods for core excitations. We report high resolution experimental measurements performed at the SOLEIL synchrotron facility. These are compared with theoretical spectra calculated using coupled cluster theory and algebraic diagrammatic construction theory. The coupled cluster singles and doubles with perturbative triples model known as CC3 is shown to accurately reproduce the experimental excitation energies as well as the spacing of the vibrational transitions. The computational results are also shown to be systematically improved within the coupled cluster hierarchy, with the coupled cluster singles, doubles, triples, and quadruples method faithfully reproducing the experimental vibrational structure.

Empirical entropic contributions in computational docking: evaluation in APS reductase complexes.

PubMed

Chang, Max W; Belew, Richard K; Carroll, Kate S; Olson, Arthur J; Goodsell, David S

2008-08-01

The results from reiterated docking experiments may be used to evaluate an empirical vibrational entropy of binding in ligand-protein complexes. We have tested several methods for evaluating the vibrational contribution to binding of 22 nucleotide analogues to the enzyme APS reductase. These include two cluster size methods that measure the probability of finding a particular conformation, a method that estimates the extent of the local energetic well by looking at the scatter of conformations within clustered results, and an RMSD-based method that uses the overall scatter and clustering of all conformations. We have also directly characterized the local energy landscape by randomly sampling around docked conformations. The simple cluster size method shows the best performance, improving the identification of correct conformations in multiple docking experiments. 2008 Wiley Periodicals, Inc.

High-performance dynamic quantum clustering on graphics processors

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

Wittek, Peter, E-mail: peterwittek@acm.org

2013-01-15

Clustering methods in machine learning may benefit from borrowing metaphors from physics. Dynamic quantum clustering associates a Gaussian wave packet with the multidimensional data points and regards them as eigenfunctions of the Schroedinger equation. The clustering structure emerges by letting the system evolve and the visual nature of the algorithm has been shown to be useful in a range of applications. Furthermore, the method only uses matrix operations, which readily lend themselves to parallelization. In this paper, we develop an implementation on graphics hardware and investigate how this approach can accelerate the computations. We achieve a speedup of up tomore » two magnitudes over a multicore CPU implementation, which proves that quantum-like methods and acceleration by graphics processing units have a great relevance to machine learning.« less

CFD Based Computations of Flexible Helicopter Blades for Stability Analysis

NASA Technical Reports Server (NTRS)

Guruswamy, Guru P.

2011-01-01

As a collaborative effort among government aerospace research laboratories an advanced version of a widely used computational fluid dynamics code, OVERFLOW, was recently released. This latest version includes additions to model flexible rotating multiple blades. In this paper, the OVERFLOW code is applied to improve the accuracy of airload computations from the linear lifting line theory that uses displacements from beam model. Data transfers required at every revolution are managed through a Unix based script that runs jobs on large super-cluster computers. Results are demonstrated for the 4-bladed UH-60A helicopter. Deviations of computed data from flight data are evaluated. Fourier analysis post-processing that is suitable for aeroelastic stability computations are performed.

CO2 adsorption on gas-phase Cu4-xPtx (x = 0-4) clusters: a DFT study.

PubMed

Gálvez-González, Luis E; Juárez-Sánchez, J Octavio; Pacheco-Contreras, Rafael; Garzón, Ignacio L; Paz-Borbón, Lauro Oliver; Posada-Amarillas, Alvaro

2018-06-13

Transition and noble metal clusters have proven to be critical novel materials, potentially offering major advantages over conventional catalysts in a range of value-added catalytic processess such as carbon dioxide transformation to methanol. In this work, a systematic computational study of CO2 adsorption on gas-phase Cu4-xPtx (x = 0-4) clusters is performed. An exhaustive potential energy surface exploration is initially performed using our recent density functional theory basin-hopping global optimization implementation. Ground-state and low-lying energy isomers are identified for Cu4-xPtx clusters. Secondly, a CO2 molecule adsorption process is analyzed on the ground-state Cu4-xPtx configurations, as a function of cluster composition. Our results show that the gas-phase linear CO2 molecule is deformed upon adsorption, with its bend angle varying from about 132° to 139°. Cu4-xPtx cluster geometries remain unchanged after CO2 adsorption, with the exception of Cu3Pt1 and Pt4 clusters. For these particular cases, a structural conversion between the ground-state geometry and the corresponding first isomer configurations is found to be assisted by the CO2 adsorption. For all clusters, the energy barriers between the ground-state and first isomer structures are explored. Our calculated CO2 adsorption energies are found to be larger for Pt-rich clusters, exhibiting a volcano-type plot. The overall effect of a hybrid functional including dispersion forces is also discussed.

Building a Terabyte Memory Bandwidth Compute Node with Four Consumer Electronics GPUs

NASA Astrophysics Data System (ADS)

Omlin, Samuel; Räss, Ludovic; Podladchikov, Yuri

2014-05-01

GPUs released for consumer electronics are generally built with the same chip architectures as the GPUs released for professional usage. With regards to scientific computing, there are no obvious important differences in functionality or performance between the two types of releases, yet the price can differ up to one order of magnitude. For example, the consumer electronics release of the most recent NVIDIA Kepler architecture (GK110), named GeForce GTX TITAN, performed equally well in conducted memory bandwidth tests as the professional release, named Tesla K20; the consumer electronics release costs about one third of the professional release. We explain how to design and assemble a well adjusted computer with four high-end consumer electronics GPUs (GeForce GTX TITAN) combining more than 1 terabyte/s memory bandwidth. We compare the system's performance and precision with the one of hardware released for professional usage. The system can be used as a powerful workstation for scientific computing or as a compute node in a home-built GPU cluster.

Design and implementation of a hybrid MPI-CUDA model for the Smith-Waterman algorithm.

PubMed

Khaled, Heba; Faheem, Hossam El Deen Mostafa; El Gohary, Rania

2015-01-01

This paper provides a novel hybrid model for solving the multiple pair-wise sequence alignment problem combining message passing interface and CUDA, the parallel computing platform and programming model invented by NVIDIA. The proposed model targets homogeneous cluster nodes equipped with similar Graphical Processing Unit (GPU) cards. The model consists of the Master Node Dispatcher (MND) and the Worker GPU Nodes (WGN). The MND distributes the workload among the cluster working nodes and then aggregates the results. The WGN performs the multiple pair-wise sequence alignments using the Smith-Waterman algorithm. We also propose a modified implementation to the Smith-Waterman algorithm based on computing the alignment matrices row-wise. The experimental results demonstrate a considerable reduction in the running time by increasing the number of the working GPU nodes. The proposed model achieved a performance of about 12 Giga cell updates per second when we tested against the SWISS-PROT protein knowledge base running on four nodes.

Power System Decomposition for Practical Implementation of Bulk-Grid Voltage Control Methods

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

Vallem, Mallikarjuna R.; Vyakaranam, Bharat GNVSR; Holzer, Jesse T.

Power system algorithms such as AC optimal power flow and coordinated volt/var control of the bulk power system are computationally intensive and become difficult to solve in operational time frames. The computational time required to run these algorithms increases exponentially as the size of the power system increases. The solution time for multiple subsystems is less than that for solving the entire system simultaneously, and the local nature of the voltage problem lends itself to such decomposition. This paper describes an algorithm that can be used to perform power system decomposition from the point of view of the voltage controlmore » problem. Our approach takes advantage of the dominant localized effect of voltage control and is based on clustering buses according to the electrical distances between them. One of the contributions of the paper is to use multidimensional scaling to compute n-dimensional Euclidean coordinates for each bus based on electrical distance to perform algorithms like K-means clustering. A simple coordinated reactive power control of photovoltaic inverters for voltage regulation is used to demonstrate the effectiveness of the proposed decomposition algorithm and its components. The proposed decomposition method is demonstrated on the IEEE 118-bus system.« less

Computational Approaches to Simulation and Optimization of Global Aircraft Trajectories

NASA Technical Reports Server (NTRS)

Ng, Hok Kwan; Sridhar, Banavar

2016-01-01

This study examines three possible approaches to improving the speed in generating wind-optimal routes for air traffic at the national or global level. They are: (a) using the resources of a supercomputer, (b) running the computations on multiple commercially available computers and (c) implementing those same algorithms into NASAs Future ATM Concepts Evaluation Tool (FACET) and compares those to a standard implementation run on a single CPU. Wind-optimal aircraft trajectories are computed using global air traffic schedules. The run time and wait time on the supercomputer for trajectory optimization using various numbers of CPUs ranging from 80 to 10,240 units are compared with the total computational time for running the same computation on a single desktop computer and on multiple commercially available computers for potential computational enhancement through parallel processing on the computer clusters. This study also re-implements the trajectory optimization algorithm for further reduction of computational time through algorithm modifications and integrates that with FACET to facilitate the use of the new features which calculate time-optimal routes between worldwide airport pairs in a wind field for use with existing FACET applications. The implementations of trajectory optimization algorithms use MATLAB, Python, and Java programming languages. The performance evaluations are done by comparing their computational efficiencies and based on the potential application of optimized trajectories. The paper shows that in the absence of special privileges on a supercomputer, a cluster of commercially available computers provides a feasible approach for national and global air traffic system studies.

Coherent Image Layout using an Adaptive Visual Vocabulary

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

Dillard, Scott E.; Henry, Michael J.; Bohn, Shawn J.

When querying a huge image database containing millions of images, the result of the query may still contain many thousands of images that need to be presented to the user. We consider the problem of arranging such a large set of images into a visually coherent layout, one that places similar images next to each other. Image similarity is determined using a bag-of-features model, and the layout is constructed from a hierarchical clustering of the image set by mapping an in-order traversal of the hierarchy tree into a space-filling curve. This layout method provides strong locality guarantees so we aremore » able to quantitatively evaluate performance using standard image retrieval benchmarks. Performance of the bag-of-features method is best when the vocabulary is learned on the image set being clustered. Because learning a large, discriminative vocabulary is a computationally demanding task, we present a novel method for efficiently adapting a generic visual vocabulary to a particular dataset. We evaluate our clustering and vocabulary adaptation methods on a variety of image datasets and show that adapting a generic vocabulary to a particular set of images improves performance on both hierarchical clustering and image retrieval tasks.« less

Hydrogen storage in N- and B-doped graphene decorated by small platinum clusters: A computational study

NASA Astrophysics Data System (ADS)

Chen, I.-Nan; Wu, Shiuan-Yau; Chen, Hsin-Tsung

2018-05-01

In this work, we perform density functional theory (DFT) calculations to investigate the hydrogen adsorption on Pt4 cluster supported on pristine, B-, and N-doped graphene sheets. It is found that the doping B or N atom in the graphene could enhance the interaction between the Pt4 cluster and the supporting substrate. The first H2 molecule is found to be dissociative chemisorption on the three substrates. Further, dissociative and molecular adsorption of multiple H2 molecules are co-adsorbed on the three substrates. In addition, the interaction between Pt4(H2)x and the substrate is illustrated for the stability of Pt4(H2)x on the substrate. AIMD simulation is also performed to verify the stability and hydrogen storage. Accordingly, the B-graphene is predicted to be the most potential materials for hydrogen storage among these three materials.

Use of advanced particle methods in modeling space propulsion and its supersonic expansions

NASA Astrophysics Data System (ADS)

Borner, Arnaud

This research discusses the use of advanced kinetic particle methods such as Molecular Dynamics (MD) and direct simulation Monte Carlo (DSMC) to model space propulsion systems such as electrospray thrusters and their supersonic expansions. MD simulations are performed to model an electrospray thruster for the ionic liquid (IL) EMIM--BF4 using coarse-grained (CG) potentials. The model is initially featuring a constant electric field applied in the longitudinal direction. Two coarse-grained potentials are compared, and the effective-force CG (EFCG) potential is found to predict the formation of the Taylor cone, the cone-jet, and other extrusion modes for similar electric fields and mass flow rates observed in experiments of a IL fed capillary-tip-extractor system better than the simple CG potential. Later, one-dimensional and fully transient three-dimensional electric fields, the latter solving Poisson's equation to take into account the electric field due to space charge at each timestep, are computed by coupling the MD model to a Poisson solver. It is found that the inhomogeneous electric field as well as that of the IL space-charge improve agreement between modeling and experiment. The boundary conditions (BCs) are found to have a substantial impact on the potential and electric field, and the tip BC is introduced and compared to the two previous BCs, named plate and needle, showing good improvement by reducing unrealistically high radial electric fields generated in the vicinity of the capillary tip. The influence of the different boundary condition models on charged species currents as a function of the mass flow rate is studied, and it is found that a constant electric field model gives similar agreement to the more rigorous and computationally expensive tip boundary condition at lower flow rates. However, at higher mass flow rates the MD simulations with the constant electric field produces extruded particles with higher Coulomb energy per ion, consistent with droplet formation. Supersonic expansions to vacuum produce clusters of sufficiently small size that properties such as heat capacities and latent heat of evaporation cannot be described by bulk vapor thermodynamic values. Therefore, MD simulations are performed to compute the evaporation rate of small water clusters as a function of temperature and size and the rates are found to agree with Unimolecular Dissociation Theory (UDT) and Classical Nucleation Theory (CNT). The heat capacities and latent heat of vaporization obtained from Monte-Carlo Canonical-Ensemble (MCCE) simulations are used in DSMC simulations of two experiments that measured Rayleigh scattering and terminal dimer mole fraction of supersonic water-jet expansions. Water-cluster temperature and size are found to be influenced by the use of kinetic rather than thermodynamic heat-capacity and latent-heat values as well as the nucleation model. Additionally, MD simulations of water condensation in a one-dimensional free expansion are performed to simulate the conditions in the core of a plume. We find that the internal structure of the clusters formed depends on the stagnation temperature conditions. Clusters of sizes 21 and 324 are studied in detail, and their radial distribution functions (RDF) are computed and compared to reported RDFs for solid amorphous ice clusters. Dielectric properties of liquid water and water clusters are investigated, and the static dielectric constant, dipole moment autocorrelation function and relative permittivity are computed by means of MD simulations.

Development of a SaaS application probe to the physical properties of the Earth's interior: An attempt at moving HPC to the cloud

NASA Astrophysics Data System (ADS)

Huang, Qian

2014-09-01

Scientific computing often requires the availability of a massive number of computers for performing large-scale simulations, and computing in mineral physics is no exception. In order to investigate physical properties of minerals at extreme conditions in computational mineral physics, parallel computing technology is used to speed up the performance by utilizing multiple computer resources to process a computational task simultaneously thereby greatly reducing computation time. Traditionally, parallel computing has been addressed by using High Performance Computing (HPC) solutions and installed facilities such as clusters and super computers. Today, it has been seen that there is a tremendous growth in cloud computing. Infrastructure as a Service (IaaS), the on-demand and pay-as-you-go model, creates a flexible and cost-effective mean to access computing resources. In this paper, a feasibility report of HPC on a cloud infrastructure is presented. It is found that current cloud services in IaaS layer still need to improve performance to be useful to research projects. On the other hand, Software as a Service (SaaS), another type of cloud computing, is introduced into an HPC system for computing in mineral physics, and an application of which is developed. In this paper, an overall description of this SaaS application is presented. This contribution can promote cloud application development in computational mineral physics, and cross-disciplinary studies.

Deep Supervised, but Not Unsupervised, Models May Explain IT Cortical Representation

PubMed Central

Khaligh-Razavi, Seyed-Mahdi; Kriegeskorte, Nikolaus

2014-01-01

Inferior temporal (IT) cortex in human and nonhuman primates serves visual object recognition. Computational object-vision models, although continually improving, do not yet reach human performance. It is unclear to what extent the internal representations of computational models can explain the IT representation. Here we investigate a wide range of computational model representations (37 in total), testing their categorization performance and their ability to account for the IT representational geometry. The models include well-known neuroscientific object-recognition models (e.g. HMAX, VisNet) along with several models from computer vision (e.g. SIFT, GIST, self-similarity features, and a deep convolutional neural network). We compared the representational dissimilarity matrices (RDMs) of the model representations with the RDMs obtained from human IT (measured with fMRI) and monkey IT (measured with cell recording) for the same set of stimuli (not used in training the models). Better performing models were more similar to IT in that they showed greater clustering of representational patterns by category. In addition, better performing models also more strongly resembled IT in terms of their within-category representational dissimilarities. Representational geometries were significantly correlated between IT and many of the models. However, the categorical clustering observed in IT was largely unexplained by the unsupervised models. The deep convolutional network, which was trained by supervision with over a million category-labeled images, reached the highest categorization performance and also best explained IT, although it did not fully explain the IT data. Combining the features of this model with appropriate weights and adding linear combinations that maximize the margin between animate and inanimate objects and between faces and other objects yielded a representation that fully explained our IT data. Overall, our results suggest that explaining IT requires computational features trained through supervised learning to emphasize the behaviorally important categorical divisions prominently reflected in IT. PMID:25375136

Scheduling Operations for Massive Heterogeneous Clusters

NASA Technical Reports Server (NTRS)

Humphrey, John; Spagnoli, Kyle

2013-01-01

High-performance computing (HPC) programming has become increasingly difficult with the advent of hybrid supercomputers consisting of multicore CPUs and accelerator boards such as the GPU. Manual tuning of software to achieve high performance on this type of machine has been performed by programmers. This is needlessly difficult and prone to being invalidated by new hardware, new software, or changes in the underlying code. A system was developed for task-based representation of programs, which when coupled with a scheduler and runtime system, allows for many benefits, including higher performance and utilization of computational resources, easier programming and porting, and adaptations of code during runtime. The system consists of a method of representing computer algorithms as a series of data-dependent tasks. The series forms a graph, which can be scheduled for execution on many nodes of a supercomputer efficiently by a computer algorithm. The schedule is executed by a dispatch component, which is tailored to understand all of the hardware types that may be available within the system. The scheduler is informed by a cluster mapping tool, which generates a topology of available resources and their strengths and communication costs. Software is decoupled from its hardware, which aids in porting to future architectures. A computer algorithm schedules all operations, which for systems of high complexity (i.e., most NASA codes), cannot be performed optimally by a human. The system aids in reducing repetitive code, such as communication code, and aids in the reduction of redundant code across projects. It adds new features to code automatically, such as recovering from a lost node or the ability to modify the code while running. In this project, the innovators at the time of this reporting intend to develop two distinct technologies that build upon each other and both of which serve as building blocks for more efficient HPC usage. First is the scheduling and dynamic execution framework, and the second is scalable linear algebra libraries that are built directly on the former.

A convergent model for distributed processing of Big Sensor Data in urban engineering networks

NASA Astrophysics Data System (ADS)

Parygin, D. S.; Finogeev, A. G.; Kamaev, V. A.; Finogeev, A. A.; Gnedkova, E. P.; Tyukov, A. P.

2017-01-01

The problems of development and research of a convergent model of the grid, cloud, fog and mobile computing for analytical Big Sensor Data processing are reviewed. The model is meant to create monitoring systems of spatially distributed objects of urban engineering networks and processes. The proposed approach is the convergence model of the distributed data processing organization. The fog computing model is used for the processing and aggregation of sensor data at the network nodes and/or industrial controllers. The program agents are loaded to perform computing tasks for the primary processing and data aggregation. The grid and the cloud computing models are used for integral indicators mining and accumulating. A computing cluster has a three-tier architecture, which includes the main server at the first level, a cluster of SCADA system servers at the second level, a lot of GPU video cards with the support for the Compute Unified Device Architecture at the third level. The mobile computing model is applied to visualize the results of intellectual analysis with the elements of augmented reality and geo-information technologies. The integrated indicators are transferred to the data center for accumulation in a multidimensional storage for the purpose of data mining and knowledge gaining.

Parallel Evolutionary Optimization for Neuromorphic Network Training

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

Schuman, Catherine D; Disney, Adam; Singh, Susheela

One of the key impediments to the success of current neuromorphic computing architectures is the issue of how best to program them. Evolutionary optimization (EO) is one promising programming technique; in particular, its wide applicability makes it especially attractive for neuromorphic architectures, which can have many different characteristics. In this paper, we explore different facets of EO on a spiking neuromorphic computing model called DANNA. We focus on the performance of EO in the design of our DANNA simulator, and on how to structure EO on both multicore and massively parallel computing systems. We evaluate how our parallel methods impactmore » the performance of EO on Titan, the U.S.'s largest open science supercomputer, and BOB, a Beowulf-style cluster of Raspberry Pi's. We also focus on how to improve the EO by evaluating commonality in higher performing neural networks, and present the result of a study that evaluates the EO performed by Titan.« less

Computer-aided detection and diagnosis of masses and clustered microcalcifications from digital mammograms

NASA Astrophysics Data System (ADS)

Nishikawa, Robert M.; Giger, Maryellen L.; Doi, Kunio; Vyborny, Carl J.; Schmidt, Robert A.; Metz, Charles E.; Wu, Chris Y.; Yin, Fang-Fang; Jiang, Yulei; Huo, Zhimin; Lu, Ping; Zhang, Wei; Ema, Takahiro; Bick, Ulrich; Papaioannou, John; Nagel, Rufus H.

1993-07-01

We are developing an 'intelligent' workstation to assist radiologists in diagnosing breast cancer from mammograms. The hardware for the workstation will consist of a film digitizer, a high speed computer, a large volume storage device, a film printer, and 4 high resolution CRT monitors. The software for the workstation is a comprehensive package of automated detection and classification schemes. Two rule-based detection schemes have been developed, one for breast masses and the other for clustered microcalcifications. The sensitivity of both schemes is 85% with a false-positive rate of approximately 3.0 and 1.5 false detections per image, for the mass and cluster detection schemes, respectively. Computerized classification is performed by an artificial neural network (ANN). The ANN has a sensitivity of 100% with a specificity of 60%. Currently, the ANN, which is a three-layer, feed-forward network, requires as input ratings of 14 different radiographic features of the mammogram that were determined subjectively by a radiologist. We are in the process of developing automated techniques to objectively determine these 14 features. The workstation will be placed in the clinical reading area of the radiology department in the near future, where controlled clinical tests will be performed to measure its efficacy.

False Discovery Control in Large-Scale Spatial Multiple Testing

PubMed Central

Sun, Wenguang; Reich, Brian J.; Cai, T. Tony; Guindani, Michele; Schwartzman, Armin

2014-01-01

Summary This article develops a unified theoretical and computational framework for false discovery control in multiple testing of spatial signals. We consider both point-wise and cluster-wise spatial analyses, and derive oracle procedures which optimally control the false discovery rate, false discovery exceedance and false cluster rate, respectively. A data-driven finite approximation strategy is developed to mimic the oracle procedures on a continuous spatial domain. Our multiple testing procedures are asymptotically valid and can be effectively implemented using Bayesian computational algorithms for analysis of large spatial data sets. Numerical results show that the proposed procedures lead to more accurate error control and better power performance than conventional methods. We demonstrate our methods for analyzing the time trends in tropospheric ozone in eastern US. PMID:25642138

Data Characterization Using Artificial-Star Tests: Performance Evaluation

NASA Astrophysics Data System (ADS)

Hu, Yi; Deng, Licai; de Grijs, Richard; Liu, Qiang

2011-01-01

Traditional artificial-star tests are widely applied to photometry in crowded stellar fields. However, to obtain reliable binary fractions (and their uncertainties) of remote, dense, and rich star clusters, one needs to recover huge numbers of artificial stars. Hence, this will consume much computation time for data reduction of the images to which the artificial stars must be added. In this article, we present a new method applicable to data sets characterized by stable, well-defined, point-spread functions, in which we add artificial stars to the retrieved-data catalog instead of to the raw images. Taking the young Large Magellanic Cloud cluster NGC 1818 as an example, we compare results from both methods and show that they are equivalent, while our new method saves significant computational time.

Hierarchically clustered adaptive quantization CMAC and its learning convergence.

PubMed

Teddy, S D; Lai, E M K; Quek, C

2007-11-01

The cerebellar model articulation controller (CMAC) neural network (NN) is a well-established computational model of the human cerebellum. Nevertheless, there are two major drawbacks associated with the uniform quantization scheme of the CMAC network. They are the following: (1) a constant output resolution associated with the entire input space and (2) the generalization-accuracy dilemma. Moreover, the size of the CMAC network is an exponential function of the number of inputs. Depending on the characteristics of the training data, only a small percentage of the entire set of CMAC memory cells is utilized. Therefore, the efficient utilization of the CMAC memory is a crucial issue. One approach is to quantize the input space nonuniformly. For existing nonuniformly quantized CMAC systems, there is a tradeoff between memory efficiency and computational complexity. Inspired by the underlying organizational mechanism of the human brain, this paper presents a novel CMAC architecture named hierarchically clustered adaptive quantization CMAC (HCAQ-CMAC). HCAQ-CMAC employs hierarchical clustering for the nonuniform quantization of the input space to identify significant input segments and subsequently allocating more memory cells to these regions. The stability of the HCAQ-CMAC network is theoretically guaranteed by the proof of its learning convergence. The performance of the proposed network is subsequently benchmarked against the original CMAC network, as well as two other existing CMAC variants on two real-life applications, namely, automated control of car maneuver and modeling of the human blood glucose dynamics. The experimental results have demonstrated that the HCAQ-CMAC network offers an efficient memory allocation scheme and improves the generalization and accuracy of the network output to achieve better or comparable performances with smaller memory usages. Index Terms-Cerebellar model articulation controller (CMAC), hierarchical clustering, hierarchically clustered adaptive quantization CMAC (HCAQ-CMAC), learning convergence, nonuniform quantization.

A clustering-based graph Laplacian framework for value function approximation in reinforcement learning.

PubMed

Xu, Xin; Huang, Zhenhua; Graves, Daniel; Pedrycz, Witold

2014-12-01

In order to deal with the sequential decision problems with large or continuous state spaces, feature representation and function approximation have been a major research topic in reinforcement learning (RL). In this paper, a clustering-based graph Laplacian framework is presented for feature representation and value function approximation (VFA) in RL. By making use of clustering-based techniques, that is, K-means clustering or fuzzy C-means clustering, a graph Laplacian is constructed by subsampling in Markov decision processes (MDPs) with continuous state spaces. The basis functions for VFA can be automatically generated from spectral analysis of the graph Laplacian. The clustering-based graph Laplacian is integrated with a class of approximation policy iteration algorithms called representation policy iteration (RPI) for RL in MDPs with continuous state spaces. Simulation and experimental results show that, compared with previous RPI methods, the proposed approach needs fewer sample points to compute an efficient set of basis functions and the learning control performance can be improved for a variety of parameter settings.

Hierarchical Adaptive Means (HAM) clustering for hardware-efficient, unsupervised and real-time spike sorting.

PubMed

Paraskevopoulou, Sivylla E; Wu, Di; Eftekhar, Amir; Constandinou, Timothy G

2014-09-30

This work presents a novel unsupervised algorithm for real-time adaptive clustering of neural spike data (spike sorting). The proposed Hierarchical Adaptive Means (HAM) clustering method combines centroid-based clustering with hierarchical cluster connectivity to classify incoming spikes using groups of clusters. It is described how the proposed method can adaptively track the incoming spike data without requiring any past history, iteration or training and autonomously determines the number of spike classes. Its performance (classification accuracy) has been tested using multiple datasets (both simulated and recorded) achieving a near-identical accuracy compared to k-means (using 10-iterations and provided with the number of spike classes). Also, its robustness in applying to different feature extraction methods has been demonstrated by achieving classification accuracies above 80% across multiple datasets. Last but crucially, its low complexity, that has been quantified through both memory and computation requirements makes this method hugely attractive for future hardware implementation. Copyright © 2014 Elsevier B.V. All rights reserved.

Modeling of Cluster-Induced Turbulence in Particle-Laden Channel Flow

NASA Astrophysics Data System (ADS)

Baker, Michael; Capecelatro, Jesse; Kong, Bo; Fox, Rodney; Desjardins, Olivier

2017-11-01

A phenomenon often observed in gas-solid flows is the formation of mesoscale clusters of particles due to the relative motion between the solid and fluid phases that is sustained through the dampening of collisional particle motion from interphase momentum coupling inside these clusters. The formation of such sustained clusters, leading to cluster-induced turbulence (CIT), can have a significant impact in industrial processes, particularly in regards to mixing, reaction progress, and heat transfer. Both Euler-Lagrange (EL) and Euler-Euler anisotropic Gaussian (EE-AG) approaches are used in this work to perform mesoscale simulations of CIT in fully developed gas-particle channel flow. The results from these simulations are applied in the development of a two-phase Reynolds-Averaged Navier-Stokes (RANS) model to capture the wall-normal flow characteristics in a less computationally expensive manner. Parameters such as mass loading, particle size, and gas velocity are varied to examine their respective impact on cluster formation and turbulence statistics. Acknowledging support from the NSF (AN:1437865).

Towards enhancement of performance of K-means clustering using nature-inspired optimization algorithms.

PubMed

Fong, Simon; Deb, Suash; Yang, Xin-She; Zhuang, Yan

2014-01-01

Traditional K-means clustering algorithms have the drawback of getting stuck at local optima that depend on the random values of initial centroids. Optimization algorithms have their advantages in guiding iterative computation to search for global optima while avoiding local optima. The algorithms help speed up the clustering process by converging into a global optimum early with multiple search agents in action. Inspired by nature, some contemporary optimization algorithms which include Ant, Bat, Cuckoo, Firefly, and Wolf search algorithms mimic the swarming behavior allowing them to cooperatively steer towards an optimal objective within a reasonable time. It is known that these so-called nature-inspired optimization algorithms have their own characteristics as well as pros and cons in different applications. When these algorithms are combined with K-means clustering mechanism for the sake of enhancing its clustering quality by avoiding local optima and finding global optima, the new hybrids are anticipated to produce unprecedented performance. In this paper, we report the results of our evaluation experiments on the integration of nature-inspired optimization methods into K-means algorithms. In addition to the standard evaluation metrics in evaluating clustering quality, the extended K-means algorithms that are empowered by nature-inspired optimization methods are applied on image segmentation as a case study of application scenario.

Towards Enhancement of Performance of K-Means Clustering Using Nature-Inspired Optimization Algorithms

PubMed Central

Deb, Suash; Yang, Xin-She

2014-01-01

Traditional K-means clustering algorithms have the drawback of getting stuck at local optima that depend on the random values of initial centroids. Optimization algorithms have their advantages in guiding iterative computation to search for global optima while avoiding local optima. The algorithms help speed up the clustering process by converging into a global optimum early with multiple search agents in action. Inspired by nature, some contemporary optimization algorithms which include Ant, Bat, Cuckoo, Firefly, and Wolf search algorithms mimic the swarming behavior allowing them to cooperatively steer towards an optimal objective within a reasonable time. It is known that these so-called nature-inspired optimization algorithms have their own characteristics as well as pros and cons in different applications. When these algorithms are combined with K-means clustering mechanism for the sake of enhancing its clustering quality by avoiding local optima and finding global optima, the new hybrids are anticipated to produce unprecedented performance. In this paper, we report the results of our evaluation experiments on the integration of nature-inspired optimization methods into K-means algorithms. In addition to the standard evaluation metrics in evaluating clustering quality, the extended K-means algorithms that are empowered by nature-inspired optimization methods are applied on image segmentation as a case study of application scenario. PMID:25202730

Performance comparison of independent component analysis algorithms for fetal cardiac signal reconstruction: a study on synthetic fMCG data

NASA Astrophysics Data System (ADS)

Mantini, D.; Hild, K. E., II; Alleva, G.; Comani, S.

2006-02-01

Independent component analysis (ICA) algorithms have been successfully used for signal extraction tasks in the field of biomedical signal processing. We studied the performances of six algorithms (FastICA, CubICA, JADE, Infomax, TDSEP and MRMI-SIG) for fetal magnetocardiography (fMCG). Synthetic datasets were used to check the quality of the separated components against the original traces. Real fMCG recordings were simulated with linear combinations of typical fMCG source signals: maternal and fetal cardiac activity, ambient noise, maternal respiration, sensor spikes and thermal noise. Clusters of different dimensions (19, 36 and 55 sensors) were prepared to represent different MCG systems. Two types of signal-to-interference ratios (SIR) were measured. The first involves averaging over all estimated components and the second is based solely on the fetal trace. The computation time to reach a minimum of 20 dB SIR was measured for all six algorithms. No significant dependency on gestational age or cluster dimension was observed. Infomax performed poorly when a sub-Gaussian source was included; TDSEP and MRMI-SIG were sensitive to additive noise, whereas FastICA, CubICA and JADE showed the best performances. Of all six methods considered, FastICA had the best overall performance in terms of both separation quality and computation times.

Implementation of Parallel Dynamic Simulation on Shared-Memory vs. Distributed-Memory Environments

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

Jin, Shuangshuang; Chen, Yousu; Wu, Di

2015-12-09

Power system dynamic simulation computes the system response to a sequence of large disturbance, such as sudden changes in generation or load, or a network short circuit followed by protective branch switching operation. It consists of a large set of differential and algebraic equations, which is computational intensive and challenging to solve using single-processor based dynamic simulation solution. High-performance computing (HPC) based parallel computing is a very promising technology to speed up the computation and facilitate the simulation process. This paper presents two different parallel implementations of power grid dynamic simulation using Open Multi-processing (OpenMP) on shared-memory platform, and Messagemore » Passing Interface (MPI) on distributed-memory clusters, respectively. The difference of the parallel simulation algorithms and architectures of the two HPC technologies are illustrated, and their performances for running parallel dynamic simulation are compared and demonstrated.« less

A Method to Compute Periodic Sums

DTIC Science & Technology

2013-10-15

the absolute performance of the present meth- ods with the smooth particle mesh Ewald ( SPME ) and other algorithms for periodic summation due to a...can be done using published data [14] comparing perfor- mance of the SPME and FMM-type PWA implementation for clusters, for relatively small size

Dynamic clustering scheme based on the coordination of management and control in multi-layer and multi-region intelligent optical network

NASA Astrophysics Data System (ADS)

Niu, Xiaoliang; Yuan, Fen; Huang, Shanguo; Guo, Bingli; Gu, Wanyi

2011-12-01

A Dynamic clustering scheme based on coordination of management and control is proposed to reduce network congestion rate and improve the blocking performance of hierarchical routing in Multi-layer and Multi-region intelligent optical network. Its implement relies on mobile agent (MA) technology, which has the advantages of efficiency, flexibility, functional and scalability. The paper's major contribution is to adjust dynamically domain when the performance of working network isn't in ideal status. And the incorporation of centralized NMS and distributed MA control technology migrate computing process to control plane node which releases the burden of NMS and improves process efficiently. Experiments are conducted on Multi-layer and multi-region Simulation Platform for Optical Network (MSPON) to assess the performance of the scheme.

Exhaustive comparison and classification of ligand-binding surfaces in proteins

PubMed Central

Murakami, Yoichi; Kinoshita, Kengo; Kinjo, Akira R; Nakamura, Haruki

2013-01-01

Many proteins function by interacting with other small molecules (ligands). Identification of ligand-binding sites (LBS) in proteins can therefore help to infer their molecular functions. A comprehensive comparison among local structures of LBSs was previously performed, in order to understand their relationships and to classify their structural motifs. However, similar exhaustive comparison among local surfaces of LBSs (patches) has never been performed, due to computational complexity. To enhance our understanding of LBSs, it is worth performing such comparisons among patches and classifying them based on similarities of their surface configurations and electrostatic potentials. In this study, we first developed a rapid method to compare two patches. We then clustered patches corresponding to the same PDB chemical component identifier for a ligand, and selected a representative patch from each cluster. We subsequently exhaustively as compared the representative patches and clustered them using similarity score, PatSim. Finally, the resultant PatSim scores were compared with similarities of atomic structures of the LBSs and those of the ligand-binding protein sequences and functions. Consequently, we classified the patches into ∼2000 well-characterized clusters. We found that about 63% of these clusters are used in identical protein folds, although about 25% of the clusters are conserved in distantly related proteins and even in proteins with cross-fold similarity. Furthermore, we showed that patches with higher PatSim score have potential to be involved in similar biological processes. PMID:23934772

Building high-performance system for processing a daily large volume of Chinese satellites imagery

NASA Astrophysics Data System (ADS)

Deng, Huawu; Huang, Shicun; Wang, Qi; Pan, Zhiqiang; Xin, Yubin

2014-10-01

The number of Earth observation satellites from China increases dramatically recently and those satellites are acquiring a large volume of imagery daily. As the main portal of image processing and distribution from those Chinese satellites, the China Centre for Resources Satellite Data and Application (CRESDA) has been working with PCI Geomatics during the last three years to solve two issues in this regard: processing the large volume of data (about 1,500 scenes or 1 TB per day) in a timely manner and generating geometrically accurate orthorectified products. After three-year research and development, a high performance system has been built and successfully delivered. The high performance system has a service oriented architecture and can be deployed to a cluster of computers that may be configured with high end computing power. The high performance is gained through, first, making image processing algorithms into parallel computing by using high performance graphic processing unit (GPU) cards and multiple cores from multiple CPUs, and, second, distributing processing tasks to a cluster of computing nodes. While achieving up to thirty (and even more) times faster in performance compared with the traditional practice, a particular methodology was developed to improve the geometric accuracy of images acquired from Chinese satellites (including HJ-1 A/B, ZY-1-02C, ZY-3, GF-1, etc.). The methodology consists of fully automatic collection of dense ground control points (GCP) from various resources and then application of those points to improve the photogrammetric model of the images. The delivered system is up running at CRESDA for pre-operational production and has been and is generating good return on investment by eliminating a great amount of manual labor and increasing more than ten times of data throughput daily with fewer operators. Future work, such as development of more performance-optimized algorithms, robust image matching methods and application workflows, is identified to improve the system in the coming years.

Dynamic provisioning of a HEP computing infrastructure on a shared hybrid HPC system

NASA Astrophysics Data System (ADS)

Meier, Konrad; Fleig, Georg; Hauth, Thomas; Janczyk, Michael; Quast, Günter; von Suchodoletz, Dirk; Wiebelt, Bernd

2016-10-01

Experiments in high-energy physics (HEP) rely on elaborate hardware, software and computing systems to sustain the high data rates necessary to study rare physics processes. The Institut fr Experimentelle Kernphysik (EKP) at KIT is a member of the CMS and Belle II experiments, located at the LHC and the Super-KEKB accelerators, respectively. These detectors share the requirement, that enormous amounts of measurement data must be processed and analyzed and a comparable amount of simulated events is required to compare experimental results with theoretical predictions. Classical HEP computing centers are dedicated sites which support multiple experiments and have the required software pre-installed. Nowadays, funding agencies encourage research groups to participate in shared HPC cluster models, where scientist from different domains use the same hardware to increase synergies. This shared usage proves to be challenging for HEP groups, due to their specialized software setup which includes a custom OS (often Scientific Linux), libraries and applications. To overcome this hurdle, the EKP and data center team of the University of Freiburg have developed a system to enable the HEP use case on a shared HPC cluster. To achieve this, an OpenStack-based virtualization layer is installed on top of a bare-metal cluster. While other user groups can run their batch jobs via the Moab workload manager directly on bare-metal, HEP users can request virtual machines with a specialized machine image which contains a dedicated operating system and software stack. In contrast to similar installations, in this hybrid setup, no static partitioning of the cluster into a physical and virtualized segment is required. As a unique feature, the placement of the virtual machine on the cluster nodes is scheduled by Moab and the job lifetime is coupled to the lifetime of the virtual machine. This allows for a seamless integration with the jobs sent by other user groups and honors the fairshare policies of the cluster. The developed thin integration layer between OpenStack and Moab can be adapted to other batch servers and virtualization systems, making the concept also applicable for other cluster operators. This contribution will report on the concept and implementation of an OpenStack-virtualized cluster used for HEP workflows. While the full cluster will be installed in spring 2016, a test-bed setup with 800 cores has been used to study the overall system performance and dedicated HEP jobs were run in a virtualized environment over many weeks. Furthermore, the dynamic integration of the virtualized worker nodes, depending on the workload at the institute's computing system, will be described.

Transformation of OODT CAS to Perform Larger Tasks

NASA Technical Reports Server (NTRS)

Mattmann, Chris; Freeborn, Dana; Crichton, Daniel; Hughes, John; Ramirez, Paul; Hardman, Sean; Woollard, David; Kelly, Sean

2008-01-01

A computer program denoted OODT CAS has been transformed to enable performance of larger tasks that involve greatly increased data volumes and increasingly intensive processing of data on heterogeneous, geographically dispersed computers. Prior to the transformation, OODT CAS (also alternatively denoted, simply, 'CAS') [wherein 'OODT' signifies 'Object-Oriented Data Technology' and 'CAS' signifies 'Catalog and Archive Service'] was a proven software component used to manage scientific data from spaceflight missions. In the transformation, CAS was split into two separate components representing its canonical capabilities: file management and workflow management. In addition, CAS was augmented by addition of a resource-management component. This third component enables CAS to manage heterogeneous computing by use of diverse resources, including high-performance clusters of computers, commodity computing hardware, and grid computing infrastructures. CAS is now more easily maintainable, evolvable, and reusable. These components can be used separately or, taking advantage of synergies, can be used together. Other elements of the transformation included addition of a separate Web presentation layer that supports distribution of data products via Really Simple Syndication (RSS) feeds, and provision for full Resource Description Framework (RDF) exports of metadata.

Solving Coupled Gross--Pitaevskii Equations on a Cluster of PlayStation 3 Computers

NASA Astrophysics Data System (ADS)

Edwards, Mark; Heward, Jeffrey; Clark, C. W.

2009-05-01

At Georgia Southern University we have constructed an 8+1--node cluster of Sony PlayStation 3 (PS3) computers with the intention of using this computing resource to solve problems related to the behavior of ultra--cold atoms in general with a particular emphasis on studying bose--bose and bose--fermi mixtures confined in optical lattices. As a first project that uses this computing resource, we have implemented a parallel solver of the coupled time--dependent, one--dimensional Gross--Pitaevskii (TDGP) equations. These equations govern the behavior of dual-- species bosonic mixtures. We chose the split--operator/FFT to solve the coupled 1D TDGP equations. The fast Fourier transform component of this solver can be readily parallelized on the PS3 cpu known as the Cell Broadband Engine (CellBE). Each CellBE chip contains a single 64--bit PowerPC Processor Element known as the PPE and eight ``Synergistic Processor Element'' identified as the SPE's. We report on this algorithm and compare its performance to a non--parallel solver as applied to modeling evaporative cooling in dual--species bosonic mixtures.

Linear solver performance in elastoplastic problem solution on GPU cluster

NASA Astrophysics Data System (ADS)

Khalevitsky, Yu. V.; Konovalov, A. V.; Burmasheva, N. V.; Partin, A. S.

2017-12-01

Applying the finite element method to severe plastic deformation problems involves solving linear equation systems. While the solution procedure is relatively hard to parallelize and computationally intensive by itself, a long series of large scale systems need to be solved for each problem. When dealing with fine computational meshes, such as in the simulations of three-dimensional metal matrix composite microvolume deformation, tens and hundreds of hours may be needed to complete the whole solution procedure, even using modern supercomputers. In general, one of the preconditioned Krylov subspace methods is used in a linear solver for such problems. The method convergence highly depends on the operator spectrum of a problem stiffness matrix. In order to choose the appropriate method, a series of computational experiments is used. Different methods may be preferable for different computational systems for the same problem. In this paper we present experimental data obtained by solving linear equation systems from an elastoplastic problem on a GPU cluster. The data can be used to substantiate the choice of the appropriate method for a linear solver to use in severe plastic deformation simulations.

High Performance, Dependable Multiprocessor

NASA Technical Reports Server (NTRS)

Ramos, Jeremy; Samson, John R.; Troxel, Ian; Subramaniyan, Rajagopal; Jacobs, Adam; Greco, James; Cieslewski, Grzegorz; Curreri, John; Fischer, Michael; Grobelny, Eric;

From virtual clustering analysis to self-consistent clustering analysis: a mathematical study

NASA Astrophysics Data System (ADS)

Tang, Shaoqiang; Zhang, Lei; Liu, Wing Kam

2018-03-01

In this paper, we propose a new homogenization algorithm, virtual clustering analysis (VCA), as well as provide a mathematical framework for the recently proposed self-consistent clustering analysis (SCA) (Liu et al. in Comput Methods Appl Mech Eng 306:319-341, 2016). In the mathematical theory, we clarify the key assumptions and ideas of VCA and SCA, and derive the continuous and discrete Lippmann-Schwinger equations. Based on a key postulation of "once response similarly, always response similarly", clustering is performed in an offline stage by machine learning techniques (k-means and SOM), and facilitates substantial reduction of computational complexity in an online predictive stage. The clear mathematical setup allows for the first time a convergence study of clustering refinement in one space dimension. Convergence is proved rigorously, and found to be of second order from numerical investigations. Furthermore, we propose to suitably enlarge the domain in VCA, such that the boundary terms may be neglected in the Lippmann-Schwinger equation, by virtue of the Saint-Venant's principle. In contrast, they were not obtained in the original SCA paper, and we discover these terms may well be responsible for the numerical dependency on the choice of reference material property. Since VCA enhances the accuracy by overcoming the modeling error, and reduce the numerical cost by avoiding an outer loop iteration for attaining the material property consistency in SCA, its efficiency is expected even higher than the recently proposed SCA algorithm.

An Information-Theoretic-Cluster Visualization for Self-Organizing Maps.

PubMed

Brito da Silva, Leonardo Enzo; Wunsch, Donald C

2018-06-01

Improved data visualization will be a significant tool to enhance cluster analysis. In this paper, an information-theoretic-based method for cluster visualization using self-organizing maps (SOMs) is presented. The information-theoretic visualization (IT-vis) has the same structure as the unified distance matrix, but instead of depicting Euclidean distances between adjacent neurons, it displays the similarity between the distributions associated with adjacent neurons. Each SOM neuron has an associated subset of the data set whose cardinality controls the granularity of the IT-vis and with which the first- and second-order statistics are computed and used to estimate their probability density functions. These are used to calculate the similarity measure, based on Renyi's quadratic cross entropy and cross information potential (CIP). The introduced visualizations combine the low computational cost and kernel estimation properties of the representative CIP and the data structure representation of a single-linkage-based grouping algorithm to generate an enhanced SOM-based visualization. The visual quality of the IT-vis is assessed by comparing it with other visualization methods for several real-world and synthetic benchmark data sets. Thus, this paper also contains a significant literature survey. The experiments demonstrate the IT-vis cluster revealing capabilities, in which cluster boundaries are sharply captured. Additionally, the information-theoretic visualizations are used to perform clustering of the SOM. Compared with other methods, IT-vis of large SOMs yielded the best results in this paper, for which the quality of the final partitions was evaluated using external validity indices.

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

Turner, A.; Davis, A.; University of Wisconsin-Madison, Madison, WI 53706

CCFE perform Monte-Carlo transport simulations on large and complex tokamak models such as ITER. Such simulations are challenging since streaming and deep penetration effects are equally important. In order to make such simulations tractable, both variance reduction (VR) techniques and parallel computing are used. It has been found that the application of VR techniques in such models significantly reduces the efficiency of parallel computation due to 'long histories'. VR in MCNP can be accomplished using energy-dependent weight windows. The weight window represents an 'average behaviour' of particles, and large deviations in the arriving weight of a particle give rise tomore » extreme amounts of splitting being performed and a long history. When running on parallel clusters, a long history can have a detrimental effect on the parallel efficiency - if one process is computing the long history, the other CPUs complete their batch of histories and wait idle. Furthermore some long histories have been found to be effectively intractable. To combat this effect, CCFE has developed an adaptation of MCNP which dynamically adjusts the WW where a large weight deviation is encountered. The method effectively 'de-optimises' the WW, reducing the VR performance but this is offset by a significant increase in parallel efficiency. Testing with a simple geometry has shown the method does not bias the result. This 'long history method' has enabled CCFE to significantly improve the performance of MCNP calculations for ITER on parallel clusters, and will be beneficial for any geometry combining streaming and deep penetration effects. (authors)« less

An Improved Clustering Algorithm of Tunnel Monitoring Data for Cloud Computing

PubMed Central

Zhong, Luo; Tang, KunHao; Li, Lin; Yang, Guang; Ye, JingJing

2014-01-01

With the rapid development of urban construction, the number of urban tunnels is increasing and the data they produce become more and more complex. It results in the fact that the traditional clustering algorithm cannot handle the mass data of the tunnel. To solve this problem, an improved parallel clustering algorithm based on k-means has been proposed. It is a clustering algorithm using the MapReduce within cloud computing that deals with data. It not only has the advantage of being used to deal with mass data but also is more efficient. Moreover, it is able to compute the average dissimilarity degree of each cluster in order to clean the abnormal data. PMID:24982971

Big Data Clustering via Community Detection and Hyperbolic Network Embedding in IoT Applications.

PubMed

Karyotis, Vasileios; Tsitseklis, Konstantinos; Sotiropoulos, Konstantinos; Papavassiliou, Symeon

2018-04-15

In this paper, we present a novel data clustering framework for big sensory data produced by IoT applications. Based on a network representation of the relations among multi-dimensional data, data clustering is mapped to node clustering over the produced data graphs. To address the potential very large scale of such datasets/graphs that test the limits of state-of-the-art approaches, we map the problem of data clustering to a community detection one over the corresponding data graphs. Specifically, we propose a novel computational approach for enhancing the traditional Girvan-Newman (GN) community detection algorithm via hyperbolic network embedding. The data dependency graph is embedded in the hyperbolic space via Rigel embedding, allowing more efficient computation of edge-betweenness centrality needed in the GN algorithm. This allows for more efficient clustering of the nodes of the data graph in terms of modularity, without sacrificing considerable accuracy. In order to study the operation of our approach with respect to enhancing GN community detection, we employ various representative types of artificial complex networks, such as scale-free, small-world and random geometric topologies, and frequently-employed benchmark datasets for demonstrating its efficacy in terms of data clustering via community detection. Furthermore, we provide a proof-of-concept evaluation by applying the proposed framework over multi-dimensional datasets obtained from an operational smart-city/building IoT infrastructure provided by the Federated Interoperable Semantic IoT/cloud Testbeds and Applications (FIESTA-IoT) testbed federation. It is shown that the proposed framework can be indeed used for community detection/data clustering and exploited in various other IoT applications, such as performing more energy-efficient smart-city/building sensing.

Big Data Clustering via Community Detection and Hyperbolic Network Embedding in IoT Applications

PubMed Central

Sotiropoulos, Konstantinos

2018-01-01

In this paper, we present a novel data clustering framework for big sensory data produced by IoT applications. Based on a network representation of the relations among multi-dimensional data, data clustering is mapped to node clustering over the produced data graphs. To address the potential very large scale of such datasets/graphs that test the limits of state-of-the-art approaches, we map the problem of data clustering to a community detection one over the corresponding data graphs. Specifically, we propose a novel computational approach for enhancing the traditional Girvan–Newman (GN) community detection algorithm via hyperbolic network embedding. The data dependency graph is embedded in the hyperbolic space via Rigel embedding, allowing more efficient computation of edge-betweenness centrality needed in the GN algorithm. This allows for more efficient clustering of the nodes of the data graph in terms of modularity, without sacrificing considerable accuracy. In order to study the operation of our approach with respect to enhancing GN community detection, we employ various representative types of artificial complex networks, such as scale-free, small-world and random geometric topologies, and frequently-employed benchmark datasets for demonstrating its efficacy in terms of data clustering via community detection. Furthermore, we provide a proof-of-concept evaluation by applying the proposed framework over multi-dimensional datasets obtained from an operational smart-city/building IoT infrastructure provided by the Federated Interoperable Semantic IoT/cloud Testbeds and Applications (FIESTA-IoT) testbed federation. It is shown that the proposed framework can be indeed used for community detection/data clustering and exploited in various other IoT applications, such as performing more energy-efficient smart-city/building sensing. PMID:29662043

Desktop supercomputer: what can it do?

NASA Astrophysics Data System (ADS)

Bogdanov, A.; Degtyarev, A.; Korkhov, V.

2017-12-01

The paper addresses the issues of solving complex problems that require using supercomputers or multiprocessor clusters available for most researchers nowadays. Efficient distribution of high performance computing resources according to actual application needs has been a major research topic since high-performance computing (HPC) technologies became widely introduced. At the same time, comfortable and transparent access to these resources was a key user requirement. In this paper we discuss approaches to build a virtual private supercomputer available at user's desktop: a virtual computing environment tailored specifically for a target user with a particular target application. We describe and evaluate possibilities to create the virtual supercomputer based on light-weight virtualization technologies, and analyze the efficiency of our approach compared to traditional methods of HPC resource management.

Community detection in complex networks using proximate support vector clustering

NASA Astrophysics Data System (ADS)

Wang, Feifan; Zhang, Baihai; Chai, Senchun; Xia, Yuanqing

2018-03-01

Community structure, one of the most attention attracting properties in complex networks, has been a cornerstone in advances of various scientific branches. A number of tools have been involved in recent studies concentrating on the community detection algorithms. In this paper, we propose a support vector clustering method based on a proximity graph, owing to which the introduced algorithm surpasses the traditional support vector approach both in accuracy and complexity. Results of extensive experiments undertaken on computer generated networks and real world data sets illustrate competent performances in comparison with the other counterparts.

Using the cloud to speed-up calibration of watershed-scale hydrologic models (Invited)

NASA Astrophysics Data System (ADS)

Goodall, J. L.; Ercan, M. B.; Castronova, A. M.; Humphrey, M.; Beekwilder, N.; Steele, J.; Kim, I.

2013-12-01

This research focuses on using the cloud to address computational challenges associated with hydrologic modeling. One example is calibration of a watershed-scale hydrologic model, which can take days of execution time on typical computers. While parallel algorithms for model calibration exist and some researchers have used multi-core computers or clusters to run these algorithms, these solutions do not fully address the challenge because (i) calibration can still be too time consuming even on multicore personal computers and (ii) few in the community have the time and expertise needed to manage a compute cluster. Given this, another option for addressing this challenge that we are exploring through this work is the use of the cloud for speeding-up calibration of watershed-scale hydrologic models. The cloud used in this capacity provides a means for renting a specific number and type of machines for only the time needed to perform a calibration model run. The cloud allows one to precisely balance the duration of the calibration with the financial costs so that, if the budget allows, the calibration can be performed more quickly by renting more machines. Focusing specifically on the SWAT hydrologic model and a parallel version of the DDS calibration algorithm, we show significant speed-up time across a range of watershed sizes using up to 256 cores to perform a model calibration. The tool provides a simple web-based user interface and the ability to monitor the calibration job submission process during the calibration process. Finally this talk concludes with initial work to leverage the cloud for other tasks associated with hydrologic modeling including tasks related to preparing inputs for constructing place-based hydrologic models.

Commodity Cluster Computing for Remote Sensing Applications using Red Hat LINUX

NASA Technical Reports Server (NTRS)

Dorband, John

2003-01-01

Since 1994, we have been doing research at Goddard Space Flight Center on implementing a wide variety of applications on commodity based computing clusters. This talk is about these clusters and haw they are used on these applications including ones for remote sensing.

A Parallel Neuromorphic Text Recognition System and Its Implementation on a Heterogeneous High-Performance Computing Cluster

DTIC Science & Technology

2013-01-01

M. Ahmadi, and M. Shridhar, “ Handwritten Numeral Recognition with Multiple Features and Multistage Classifiers,” Proc. IEEE Int’l Symp. Circuits...ARTICLE (Post Print) 3. DATES COVERED (From - To) SEP 2011 – SEP 2013 4. TITLE AND SUBTITLE A PARALLEL NEUROMORPHIC TEXT RECOGNITION SYSTEM AND ITS...research in computational intelligence has entered a new era. In this paper, we present an HPC-based context-aware intelligent text recognition

A harmonic linear dynamical system for prominent ECG feature extraction.

PubMed

Thi, Ngoc Anh Nguyen; Yang, Hyung-Jeong; Kim, SunHee; Do, Luu Ngoc

2014-01-01

Unsupervised mining of electrocardiography (ECG) time series is a crucial task in biomedical applications. To have efficiency of the clustering results, the prominent features extracted from preprocessing analysis on multiple ECG time series need to be investigated. In this paper, a Harmonic Linear Dynamical System is applied to discover vital prominent features via mining the evolving hidden dynamics and correlations in ECG time series. The discovery of the comprehensible and interpretable features of the proposed feature extraction methodology effectively represents the accuracy and the reliability of clustering results. Particularly, the empirical evaluation results of the proposed method demonstrate the improved performance of clustering compared to the previous main stream feature extraction approaches for ECG time series clustering tasks. Furthermore, the experimental results on real-world datasets show scalability with linear computation time to the duration of the time series.

Advances in Significance Testing for Cluster Detection

NASA Astrophysics Data System (ADS)

Coleman, Deidra Andrea

Over the past two decades, much attention has been given to data driven project goals such as the Human Genome Project and the development of syndromic surveillance systems. A major component of these types of projects is analyzing the abundance of data. Detecting clusters within the data can be beneficial as it can lead to the identification of specified sequences of DNA nucleotides that are related to important biological functions or the locations of epidemics such as disease outbreaks or bioterrorism attacks. Cluster detection techniques require efficient and accurate hypothesis testing procedures. In this dissertation, we improve upon the hypothesis testing procedures for cluster detection by enhancing distributional theory and providing an alternative method for spatial cluster detection using syndromic surveillance data. In Chapter 2, we provide an efficient method to compute the exact distribution of the number and coverage of h-clumps of a collection of words. This method involves defining a Markov chain using a minimal deterministic automaton to reduce the number of states needed for computation. We allow words of the collection to contain other words of the collection making the method more general. We use our method to compute the distributions of the number and coverage of h-clumps in the Chi motif of H. influenza.. In Chapter 3, we provide an efficient algorithm to compute the exact distribution of multiple window discrete scan statistics for higher-order, multi-state Markovian sequences. This algorithm involves defining a Markov chain to efficiently keep track of probabilities needed to compute p-values of the statistic. We use our algorithm to identify cases where the available approximation does not perform well. We also use our algorithm to detect unusual clusters of made free throw shots by National Basketball Association players during the 2009-2010 regular season. In Chapter 4, we give a procedure to detect outbreaks using syndromic surveillance data while controlling the Bayesian False Discovery Rate (BFDR). The procedure entails choosing an appropriate Bayesian model that captures the spatial dependency inherent in epidemiological data and considers all days of interest, selecting a test statistic based on a chosen measure that provides the magnitude of the maximumal spatial cluster for each day, and identifying a cutoff value that controls the BFDR for rejecting the collective null hypothesis of no outbreak over a collection of days for a specified region.We use our procedure to analyze botulism-like syndrome data collected by the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT).

Learning to Share

ERIC Educational Resources Information Center

Raths, David

2010-01-01

In the tug-of-war between researchers and IT for supercomputing resources, a centralized approach can help both sides get more bang for their buck. As 2010 began, the University of Washington was preparing to launch its first shared high-performance computing cluster, a 1,500-node system called Hyak, dedicated to research activities. Like other…

Effects of Blended Instructional Models on Math Performance

ERIC Educational Resources Information Center

Bottge, Brian A.; Ma, Xin; Gassaway, Linda; Toland, Michael D.; Butler, Mark; Cho, Sun-Joo

2014-01-01

A pretest-posttest cluster-randomized trial involving 31 middle schools and 335 students with disabilities tested the effects of combining explicit and anchored instruction on fraction computation and problem solving. Results of standardized and researcher-developed tests showed that students who were taught with the blended units outscored…

Optical potential from first principles

DOE PAGES

Rotureau, J.; Danielewicz, P.; Hagen, G.; ...

2017-02-15

Here, we develop a method to construct a microscopic optical potential from chiral interactions for nucleon-nucleus scattering. The optical potential is constructed by combining the Green’s function approach with the coupled-cluster method. To deal with the poles of the Green’s function along the real energy axis we employ a Berggren basis in the complex energy plane combined with the Lanczos method. Using this approach, we perform a proof-of-principle calculation of the optical potential for the elastic neutron scattering on 16O. For the computation of the ground-state of 16O, we use the coupled-cluster method in the singles-and-doubles approximation, while for themore » A ±1 nuclei we use particle-attached/removed equation-of-motion method truncated at two-particle-one-hole and one-particle-two-hole excitations, respectively. We verify the convergence of the optical potential and scattering phase shifts with respect to the model-space size and the number of discretized complex continuum states. We also investigate the absorptive component of the optical potential (which reflects the opening of inelastic channels) by computing its imaginary volume integral and find an almost negligible absorptive component at low-energies. To shed light on this result, we computed excited states of 16O using equation-of-motion coupled-cluster method with singles-and- doubles excitations and we found no low-lying excited states below 10 MeV. Furthermore, most excited states have a dominant two-particle-two-hole component, making higher-order particle-hole excitations necessary to achieve a precise description of these core-excited states. We conclude that the reduced absorption at low-energies can be attributed to the lack of correlations coming from the low-order cluster truncation in the employed coupled-cluster method.« less

Vertebra identification using template matching modelmp and K-means clustering.

PubMed

Larhmam, Mohamed Amine; Benjelloun, Mohammed; Mahmoudi, Saïd

2014-03-01

Accurate vertebra detection and segmentation are essential steps for automating the diagnosis of spinal disorders. This study is dedicated to vertebra alignment measurement, the first step in a computer-aided diagnosis tool for cervical spine trauma. Automated vertebral segment alignment determination is a challenging task due to low contrast imaging and noise. A software tool for segmenting vertebrae and detecting subluxations has clinical significance. A robust method was developed and tested for cervical vertebra identification and segmentation that extracts parameters used for vertebra alignment measurement. Our contribution involves a novel combination of a template matching method and an unsupervised clustering algorithm. In this method, we build a geometric vertebra mean model. To achieve vertebra detection, manual selection of the region of interest is performed initially on the input image. Subsequent preprocessing is done to enhance image contrast and detect edges. Candidate vertebra localization is then carried out by using a modified generalized Hough transform (GHT). Next, an adapted cost function is used to compute local voted centers and filter boundary data. Thereafter, a K-means clustering algorithm is applied to obtain clusters distribution corresponding to the targeted vertebrae. These clusters are combined with the vote parameters to detect vertebra centers. Rigid segmentation is then carried out by using GHT parameters. Finally, cervical spine curves are extracted to measure vertebra alignment. The proposed approach was successfully applied to a set of 66 high-resolution X-ray images. Robust detection was achieved in 97.5 % of the 330 tested cervical vertebrae. An automated vertebral identification method was developed and demonstrated to be robust to noise and occlusion. This work presents a first step toward an automated computer-aided diagnosis system for cervical spine trauma detection.

Fault-tolerant measurement-based quantum computing with continuous-variable cluster states.

PubMed

Menicucci, Nicolas C

2014-03-28

A long-standing open question about Gaussian continuous-variable cluster states is whether they enable fault-tolerant measurement-based quantum computation. The answer is yes. Initial squeezing in the cluster above a threshold value of 20.5 dB ensures that errors from finite squeezing acting on encoded qubits are below the fault-tolerance threshold of known qubit-based error-correcting codes. By concatenating with one of these codes and using ancilla-based error correction, fault-tolerant measurement-based quantum computation of theoretically indefinite length is possible with finitely squeezed cluster states.

Privacy Preserving Nearest Neighbor Search

NASA Astrophysics Data System (ADS)

Shaneck, Mark; Kim, Yongdae; Kumar, Vipin

Data mining is frequently obstructed by privacy concerns. In many cases data is distributed, and bringing the data together in one place for analysis is not possible due to privacy laws (e.g. HIPAA) or policies. Privacy preserving data mining techniques have been developed to address this issue by providing mechanisms to mine the data while giving certain privacy guarantees. In this chapter we address the issue of privacy preserving nearest neighbor search, which forms the kernel of many data mining applications. To this end, we present a novel algorithm based on secure multiparty computation primitives to compute the nearest neighbors of records in horizontally distributed data. We show how this algorithm can be used in three important data mining algorithms, namely LOF outlier detection, SNN clustering, and kNN classification. We prove the security of these algorithms under the semi-honest adversarial model, and describe methods that can be used to optimize their performance. Keywords: Privacy Preserving Data Mining, Nearest Neighbor Search, Outlier Detection, Clustering, Classification, Secure Multiparty Computation

GPU computing with Kaczmarz’s and other iterative algorithms for linear systems

PubMed Central

Elble, Joseph M.; Sahinidis, Nikolaos V.; Vouzis, Panagiotis

2009-01-01

The graphics processing unit (GPU) is used to solve large linear systems derived from partial differential equations. The differential equations studied are strongly convection-dominated, of various sizes, and common to many fields, including computational fluid dynamics, heat transfer, and structural mechanics. The paper presents comparisons between GPU and CPU implementations of several well-known iterative methods, including Kaczmarz’s, Cimmino’s, component averaging, conjugate gradient normal residual (CGNR), symmetric successive overrelaxation-preconditioned conjugate gradient, and conjugate-gradient-accelerated component-averaged row projections (CARP-CG). Computations are preformed with dense as well as general banded systems. The results demonstrate that our GPU implementation outperforms CPU implementations of these algorithms, as well as previously studied parallel implementations on Linux clusters and shared memory systems. While the CGNR method had begun to fall out of favor for solving such problems, for the problems studied in this paper, the CGNR method implemented on the GPU performed better than the other methods, including a cluster implementation of the CARP-CG method. PMID:20526446

Compute Server Performance Results

NASA Technical Reports Server (NTRS)

Stockdale, I. E.; Barton, John; Woodrow, Thomas (Technical Monitor)

1994-01-01

Parallel-vector supercomputers have been the workhorses of high performance computing. As expectations of future computing needs have risen faster than projected vector supercomputer performance, much work has been done investigating the feasibility of using Massively Parallel Processor systems as supercomputers. An even more recent development is the availability of high performance workstations which have the potential, when clustered together, to replace parallel-vector systems. We present a systematic comparison of floating point performance and price-performance for various compute server systems. A suite of highly vectorized programs was run on systems including traditional vector systems such as the Cray C90, and RISC workstations such as the IBM RS/6000 590 and the SGI R8000. The C90 system delivers 460 million floating point operations per second (FLOPS), the highest single processor rate of any vendor. However, if the price-performance ration (PPR) is considered to be most important, then the IBM and SGI processors are superior to the C90 processors. Even without code tuning, the IBM and SGI PPR's of 260 and 220 FLOPS per dollar exceed the C90 PPR of 160 FLOPS per dollar when running our highly vectorized suite,

Method of identifying clusters representing statistical dependencies in multivariate data

NASA Technical Reports Server (NTRS)

Borucki, W. J.; Card, D. H.; Lyle, G. C.

1975-01-01

Approach is first to cluster and then to compute spatial boundaries for resulting clusters. Next step is to compute, from set of Monte Carlo samples obtained from scrambled data, estimates of probabilities of obtaining at least as many points within boundaries as were actually observed in original data.

Comparative phyloinformatics of virus genes at micro and macro levels in a distributed computing environment.

PubMed

Singh, Dadabhai T; Trehan, Rahul; Schmidt, Bertil; Bretschneider, Timo

2008-01-01

Preparedness for a possible global pandemic caused by viruses such as the highly pathogenic influenza A subtype H5N1 has become a global priority. In particular, it is critical to monitor the appearance of any new emerging subtypes. Comparative phyloinformatics can be used to monitor, analyze, and possibly predict the evolution of viruses. However, in order to utilize the full functionality of available analysis packages for large-scale phyloinformatics studies, a team of computer scientists, biostatisticians and virologists is needed--a requirement which cannot be fulfilled in many cases. Furthermore, the time complexities of many algorithms involved leads to prohibitive runtimes on sequential computer platforms. This has so far hindered the use of comparative phyloinformatics as a commonly applied tool in this area. In this paper the graphical-oriented workflow design system called Quascade and its efficient usage for comparative phyloinformatics are presented. In particular, we focus on how this task can be effectively performed in a distributed computing environment. As a proof of concept, the designed workflows are used for the phylogenetic analysis of neuraminidase of H5N1 isolates (micro level) and influenza viruses (macro level). The results of this paper are hence twofold. Firstly, this paper demonstrates the usefulness of a graphical user interface system to design and execute complex distributed workflows for large-scale phyloinformatics studies of virus genes. Secondly, the analysis of neuraminidase on different levels of complexity provides valuable insights of this virus's tendency for geographical based clustering in the phylogenetic tree and also shows the importance of glycan sites in its molecular evolution. The current study demonstrates the efficiency and utility of workflow systems providing a biologist friendly approach to complex biological dataset analysis using high performance computing. In particular, the utility of the platform Quascade for deploying distributed and parallelized versions of a variety of computationally intensive phylogenetic algorithms has been shown. Secondly, the analysis of the utilized H5N1 neuraminidase datasets at macro and micro levels has clearly indicated a pattern of spatial clustering of the H5N1 viral isolates based on geographical distribution rather than temporal or host range based clustering.

The FOSS GIS Workbench on the GFZ Load Sharing Facility compute cluster

NASA Astrophysics Data System (ADS)

Löwe, P.; Klump, J.; Thaler, J.

2012-04-01

Compute clusters can be used as GIS workbenches, their wealth of resources allow us to take on geocomputation tasks which exceed the limitations of smaller systems. To harness these capabilities requires a Geographic Information System (GIS), able to utilize the available cluster configuration/architecture and a sufficient degree of user friendliness to allow for wide application. In this paper we report on the first successful porting of GRASS GIS, the oldest and largest Free Open Source (FOSS) GIS project, onto a compute cluster using Platform Computing's Load Sharing Facility (LSF). In 2008, GRASS6.3 was installed on the GFZ compute cluster, which at that time comprised 32 nodes. The interaction with the GIS was limited to the command line interface, which required further development to encapsulate the GRASS GIS business layer to facilitate its use by users not familiar with GRASS GIS. During the summer of 2011, multiple versions of GRASS GIS (v 6.4, 6.5 and 7.0) were installed on the upgraded GFZ compute cluster, now consisting of 234 nodes with 480 CPUs providing 3084 cores. The GFZ compute cluster currently offers 19 different processing queues with varying hardware capabilities and priorities, allowing for fine-grained scheduling and load balancing. After successful testing of core GIS functionalities, including the graphical user interface, mechanisms were developed to deploy scripted geocomputation tasks onto dedicated processing queues. The mechanisms are based on earlier work by NETELER et al. (2008). A first application of the new GIS functionality was the generation of maps of simulated tsunamis in the Mediterranean Sea for the Tsunami Atlas of the FP-7 TRIDEC Project (www.tridec-online.eu). For this, up to 500 processing nodes were used in parallel. Further trials included the processing of geometrically complex problems, requiring significant amounts of processing time. The GIS cluster successfully completed all these tasks, with processing times lasting up to full 20 CPU days. The deployment of GRASS GIS on a compute cluster allows our users to tackle GIS tasks previously out of reach of single workstations. In addition, this GRASS GIS cluster implementation will be made available to other users at GFZ in the course of 2012. It will thus become a research utility in the sense of "Software as a Service" (SaaS) and can be seen as our first step towards building a GFZ corporate cloud service.

Theoretical and Empirical Comparison of Big Data Image Processing with Apache Hadoop and Sun Grid Engine.

PubMed

Bao, Shunxing; Weitendorf, Frederick D; Plassard, Andrew J; Huo, Yuankai; Gokhale, Aniruddha; Landman, Bennett A

2017-02-11

The field of big data is generally concerned with the scale of processing at which traditional computational paradigms break down. In medical imaging, traditional large scale processing uses a cluster computer that combines a group of workstation nodes into a functional unit that is controlled by a job scheduler. Typically, a shared-storage network file system (NFS) is used to host imaging data. However, data transfer from storage to processing nodes can saturate network bandwidth when data is frequently uploaded/retrieved from the NFS, e.g., "short" processing times and/or "large" datasets. Recently, an alternative approach using Hadoop and HBase was presented for medical imaging to enable co-location of data storage and computation while minimizing data transfer. The benefits of using such a framework must be formally evaluated against a traditional approach to characterize the point at which simply "large scale" processing transitions into "big data" and necessitates alternative computational frameworks. The proposed Hadoop system was implemented on a production lab-cluster alongside a standard Sun Grid Engine (SGE). Theoretical models for wall-clock time and resource time for both approaches are introduced and validated. To provide real example data, three T1 image archives were retrieved from a university secure, shared web database and used to empirically assess computational performance under three configurations of cluster hardware (using 72, 109, or 209 CPU cores) with differing job lengths. Empirical results match the theoretical models. Based on these data, a comparative analysis is presented for when the Hadoop framework will be relevant and non-relevant for medical imaging.

Theoretical and empirical comparison of big data image processing with Apache Hadoop and Sun Grid Engine

NASA Astrophysics Data System (ADS)

Bao, Shunxing; Weitendorf, Frederick D.; Plassard, Andrew J.; Huo, Yuankai; Gokhale, Aniruddha; Landman, Bennett A.

2017-03-01

The field of big data is generally concerned with the scale of processing at which traditional computational paradigms break down. In medical imaging, traditional large scale processing uses a cluster computer that combines a group of workstation nodes into a functional unit that is controlled by a job scheduler. Typically, a shared-storage network file system (NFS) is used to host imaging data. However, data transfer from storage to processing nodes can saturate network bandwidth when data is frequently uploaded/retrieved from the NFS, e.g., "short" processing times and/or "large" datasets. Recently, an alternative approach using Hadoop and HBase was presented for medical imaging to enable co-location of data storage and computation while minimizing data transfer. The benefits of using such a framework must be formally evaluated against a traditional approach to characterize the point at which simply "large scale" processing transitions into "big data" and necessitates alternative computational frameworks. The proposed Hadoop system was implemented on a production lab-cluster alongside a standard Sun Grid Engine (SGE). Theoretical models for wall-clock time and resource time for both approaches are introduced and validated. To provide real example data, three T1 image archives were retrieved from a university secure, shared web database and used to empirically assess computational performance under three configurations of cluster hardware (using 72, 109, or 209 CPU cores) with differing job lengths. Empirical results match the theoretical models. Based on these data, a comparative analysis is presented for when the Hadoop framework will be relevant and nonrelevant for medical imaging.

Computational Investigation of the Geometrical and Electronic Structures of VGen-/0 (n = 1-4) Clusters by Density Functional Theory and Multiconfigurational CASSCF/CASPT2 Method.

PubMed

Tran, Van Tan; Nguyen, Minh Thao; Tran, Quoc Tri

2017-10-12

Density functional theory and the multiconfigurational CASSCF/CASPT2 method have been employed to study the low-lying states of VGe n -/0 (n = 1-4) clusters. For VGe -/0 and VGe 2 -/0 clusters, the relative energies and geometrical structures of the low-lying states are reported at the CASSCF/CASPT2 level. For the VGe 3 -/0 and VGe 4 -/0 clusters, the computational results show that due to the large contribution of the Hartree-Fock exact exchange, the hybrid B3LYP, B3PW91, and PBE0 functionals overestimate the energies of the high-spin states as compared to the pure GGA BP86 and PBE functionals and the CASPT2 method. On the basis of the pure GGA BP86 and PBE functionals and the CASSCF/CASPT2 results, the ground states of anionic and neutral clusters are defined, the relative energies of the excited states are computed, and the electron detachment energies of the anionic clusters are evaluated. The computational results are employed to give new assignments for all features in the photoelectron spectra of VGe 3 - and VGe 4 - clusters.

Next-generation genotype imputation service and methods.

PubMed

Das, Sayantan; Forer, Lukas; Schönherr, Sebastian; Sidore, Carlo; Locke, Adam E; Kwong, Alan; Vrieze, Scott I; Chew, Emily Y; Levy, Shawn; McGue, Matt; Schlessinger, David; Stambolian, Dwight; Loh, Po-Ru; Iacono, William G; Swaroop, Anand; Scott, Laura J; Cucca, Francesco; Kronenberg, Florian; Boehnke, Michael; Abecasis, Gonçalo R; Fuchsberger, Christian

2016-10-01

Genotype imputation is a key component of genetic association studies, where it increases power, facilitates meta-analysis, and aids interpretation of signals. Genotype imputation is computationally demanding and, with current tools, typically requires access to a high-performance computing cluster and to a reference panel of sequenced genomes. Here we describe improvements to imputation machinery that reduce computational requirements by more than an order of magnitude with no loss of accuracy in comparison to standard imputation tools. We also describe a new web-based service for imputation that facilitates access to new reference panels and greatly improves user experience and productivity.

Search for global-minimum geometries of medium-sized germanium clusters. II. Motif-based low-lying clusters Ge21-Ge29

NASA Astrophysics Data System (ADS)

Yoo, S.; Zeng, X. C.

2006-05-01

We performed a constrained search for the geometries of low-lying neutral germanium clusters GeN in the size range of 21⩽N⩽29. The basin-hopping global optimization method is employed for the search. The potential-energy surface is computed based on the plane-wave pseudopotential density functional theory. A new series of low-lying clusters is found on the basis of several generic structural motifs identified previously for silicon clusters [S. Yoo and X. C. Zeng, J. Chem. Phys. 124, 054304 (2006)] as well as for smaller-sized germanium clusters [S. Bulusu et al., J. Chem. Phys. 122, 164305 (2005)]. Among the generic motifs examined, we found that two motifs stand out in producing most low-lying clusters, namely, the six/nine motif, a puckered-hexagonal-ring Ge6 unit attached to a tricapped trigonal prism Ge9, and the six/ten motif, a puckered-hexagonal-ring Ge6 unit attached to a bicapped antiprism Ge10. The low-lying clusters obtained are all prolate in shape and their energies are appreciably lower than the near-spherical low-energy clusters. This result is consistent with the ion-mobility measurement in that medium-sized germanium clusters detected are all prolate in shape until the size N ˜65.

Computer-aided detection of clustered microcalcifications in multiscale bilateral filtering regularized reconstructed digital breast tomosynthesis volume

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

Samala, Ravi K., E-mail: rsamala@umich.edu; Chan, Heang-Ping; Lu, Yao

Purpose: Develop a computer-aided detection (CADe) system for clustered microcalcifications in digital breast tomosynthesis (DBT) volume enhanced with multiscale bilateral filtering (MSBF) regularization. Methods: With Institutional Review Board approval and written informed consent, two-view DBT of 154 breasts, of which 116 had biopsy-proven microcalcification (MC) clusters and 38 were free of MCs, was imaged with a General Electric GEN2 prototype DBT system. The DBT volumes were reconstructed with MSBF-regularized simultaneous algebraic reconstruction technique (SART) that was designed to enhance MCs and reduce background noise while preserving the quality of other tissue structures. The contrast-to-noise ratio (CNR) of MCs was furthermore » improved with enhancement-modulated calcification response (EMCR) preprocessing, which combined multiscale Hessian response to enhance MCs by shape and bandpass filtering to remove the low-frequency structured background. MC candidates were then located in the EMCR volume using iterative thresholding and segmented by adaptive region growing. Two sets of potential MC objects, cluster centroid objects and MC seed objects, were generated and the CNR of each object was calculated. The number of candidates in each set was controlled based on the breast volume. Dynamic clustering around the centroid objects grouped the MC candidates to form clusters. Adaptive criteria were designed to reduce false positive (FP) clusters based on the size, CNR values and the number of MCs in the cluster, cluster shape, and cluster based maximum intensity projection. Free-response receiver operating characteristic (FROC) and jackknife alternative FROC (JAFROC) analyses were used to assess the performance and compare with that of a previous study. Results: Unpaired two-tailedt-test showed a significant increase (p < 0.0001) in the ratio of CNRs for MCs with and without MSBF regularization compared to similar ratios for FPs. For view-based detection, a sensitivity of 85% was achieved at an FP rate of 2.16 per DBT volume. For case-based detection, a sensitivity of 85% was achieved at an FP rate of 0.85 per DBT volume. JAFROC analysis showed a significant improvement in the performance of the current CADe system compared to that of our previous system (p = 0.003). Conclusions: MBSF regularized SART reconstruction enhances MCs. The enhancement in the signals, in combination with properly designed adaptive threshold criteria, effective MC feature analysis, and false positive reduction techniques, leads to a significant improvement in the detection of clustered MCs in DBT.« less

Pilot-in-the-Loop CFD Method Development

DTIC Science & Technology

2017-04-20

the methods on the NAVAIR Manned Flight Simulator. Activities this period During this report period, we implemented the CRAFT CFD code on the...Penn State VLRCROE Flight simulator and performed the first Pilot-in-the-Loop PILCFD tests at Penn State using the COCOA5 clusters. The initial tests...integration of the flight simulator and Penn State computing infrastructure. Initial tests showed slower performance than real-time (3x slower than real

Differential equations as a tool for community identification.

PubMed

Krawczyk, Małgorzata J

2008-06-01

We consider the task of identification of a cluster structure in random networks. The results of two methods are presented: (i) the Newman algorithm [M. E. J. Newman and M. Girvan, Phys. Rev. E 69, 026113 (2004)]; and (ii) our method based on differential equations. A series of computer experiments is performed to check if in applying these methods we are able to determine the structure of the network. The trial networks consist initially of well-defined clusters and are disturbed by introducing noise into their connectivity matrices. Further, we show that an improvement of the previous version of our method is possible by an appropriate choice of the threshold parameter beta . With this change, the results obtained by the two methods above are similar, and our method works better, for all the computer experiments we have done.

Computational study on the molecular structures and photoelectron spectra of bimetallic oxide clusters MW2O9(-/0) (M=V, Nb, Ta).

PubMed

Chen, Wen-Jie; Zhang, Chang-Fu; Zhang, Xian-Hui; Zhang, Yong-Fan; Huang, Xin

2013-05-15

Density functional theory (DFT) and coupled cluster theory (CCSD(T)) calculations are carried out to investigate the electronic and structural properties of a series of bimetallic oxide clusters MW2O9(-/0) (M=V, Nb, Ta). Generalized Koopmans' theorem is applied to predict the vertical detachment energies (VDEs) and simulate the photoelectron spectra (PES). Theoretical calculations at the B3LYP level yield singlet and doublet ground states for the bimetallic anionic and neutral clusters, respectively. All the clusters present the six-membered ring structures with different symmetries, except that the TaW2O9(-) cluster shows a chained style with a penta-coordinated tantalum atom. Spin density analyses reveal oxygen radical species in all neutral clusters, consistent with their structural characteristics. Moreover, additional calculations are performed to study the oxidation reaction of CO molecule with the W3O9(+) cation and the isoelectronic VW2O9 cluster, and results indicate that the introduction of vanadium at tungsten site can efficiently improve the oxidation reactivity. Copyright © 2013 Elsevier B.V. All rights reserved.

Vivaldi: A Domain-Specific Language for Volume Processing and Visualization on Distributed Heterogeneous Systems.

PubMed

Choi, Hyungsuk; Choi, Woohyuk; Quan, Tran Minh; Hildebrand, David G C; Pfister, Hanspeter; Jeong, Won-Ki

2014-12-01

As the size of image data from microscopes and telescopes increases, the need for high-throughput processing and visualization of large volumetric data has become more pressing. At the same time, many-core processors and GPU accelerators are commonplace, making high-performance distributed heterogeneous computing systems affordable. However, effectively utilizing GPU clusters is difficult for novice programmers, and even experienced programmers often fail to fully leverage the computing power of new parallel architectures due to their steep learning curve and programming complexity. In this paper, we propose Vivaldi, a new domain-specific language for volume processing and visualization on distributed heterogeneous computing systems. Vivaldi's Python-like grammar and parallel processing abstractions provide flexible programming tools for non-experts to easily write high-performance parallel computing code. Vivaldi provides commonly used functions and numerical operators for customized visualization and high-throughput image processing applications. We demonstrate the performance and usability of Vivaldi on several examples ranging from volume rendering to image segmentation.

Performance of an MPI-only semiconductor device simulator on a quad socket/quad core InfiniBand platform.

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

Shadid, John Nicolas; Lin, Paul Tinphone

2009-01-01

This preliminary study considers the scaling and performance of a finite element (FE) semiconductor device simulator on a capacity cluster with 272 compute nodes based on a homogeneous multicore node architecture utilizing 16 cores. The inter-node communication backbone for this Tri-Lab Linux Capacity Cluster (TLCC) machine is comprised of an InfiniBand interconnect. The nonuniform memory access (NUMA) nodes consist of 2.2 GHz quad socket/quad core AMD Opteron processors. The performance results for this study are obtained with a FE semiconductor device simulation code (Charon) that is based on a fully-coupled Newton-Krylov solver with domain decomposition and multilevel preconditioners. Scaling andmore » multicore performance results are presented for large-scale problems of 100+ million unknowns on up to 4096 cores. A parallel scaling comparison is also presented with the Cray XT3/4 Red Storm capability platform. The results indicate that an MPI-only programming model for utilizing the multicore nodes is reasonably efficient on all 16 cores per compute node. However, the results also indicated that the multilevel preconditioner, which is critical for large-scale capability type simulations, scales better on the Red Storm machine than the TLCC machine.« less

Kappa statistic for the clustered dichotomous responses from physicians and patients

PubMed Central

Kang, Chaeryon; Qaqish, Bahjat; Monaco, Jane; Sheridan, Stacey L.; Cai, Jianwen

2013-01-01

The bootstrap method for estimating the standard error of the kappa statistic in the presence of clustered data is evaluated. Such data arise, for example, in assessing agreement between physicians and their patients regarding their understanding of the physician-patient interaction and discussions. We propose a computationally efficient procedure for generating correlated dichotomous responses for physicians and assigned patients for simulation studies. The simulation result demonstrates that the proposed bootstrap method produces better estimate of the standard error and better coverage performance compared to the asymptotic standard error estimate that ignores dependence among patients within physicians with at least a moderately large number of clusters. An example of an application to a coronary heart disease prevention study is presented. PMID:23533082

SLIC superpixels compared to state-of-the-art superpixel methods.

PubMed

Achanta, Radhakrishna; Shaji, Appu; Smith, Kevin; Lucchi, Aurelien; Fua, Pascal; Süsstrunk, Sabine

2012-11-01

Computer vision applications have come to rely increasingly on superpixels in recent years, but it is not always clear what constitutes a good superpixel algorithm. In an effort to understand the benefits and drawbacks of existing methods, we empirically compare five state-of-the-art superpixel algorithms for their ability to adhere to image boundaries, speed, memory efficiency, and their impact on segmentation performance. We then introduce a new superpixel algorithm, simple linear iterative clustering (SLIC), which adapts a k-means clustering approach to efficiently generate superpixels. Despite its simplicity, SLIC adheres to boundaries as well as or better than previous methods. At the same time, it is faster and more memory efficient, improves segmentation performance, and is straightforward to extend to supervoxel generation.

Simple, efficient allocation of modelling runs on heterogeneous clusters with MPI

USGS Publications Warehouse

Donato, David I.

2017-01-01

In scientific modelling and computation, the choice of an appropriate method for allocating tasks for parallel processing depends on the computational setting and on the nature of the computation. The allocation of independent but similar computational tasks, such as modelling runs or Monte Carlo trials, among the nodes of a heterogeneous computational cluster is a special case that has not been specifically evaluated previously. A simulation study shows that a method of on-demand (that is, worker-initiated) pulling from a bag of tasks in this case leads to reliably short makespans for computational jobs despite heterogeneity both within and between cluster nodes. A simple reference implementation in the C programming language with the Message Passing Interface (MPI) is provided.

SCEAPI: A unified Restful Web API for High-Performance Computing

NASA Astrophysics Data System (ADS)

Rongqiang, Cao; Haili, Xiao; Shasha, Lu; Yining, Zhao; Xiaoning, Wang; Xuebin, Chi

2017-10-01

The development of scientific computing is increasingly moving to collaborative web and mobile applications. All these applications need high-quality programming interface for accessing heterogeneous computing resources consisting of clusters, grid computing or cloud computing. In this paper, we introduce our high-performance computing environment that integrates computing resources from 16 HPC centers across China. Then we present a bundle of web services called SCEAPI and describe how it can be used to access HPC resources with HTTP or HTTPs protocols. We discuss SCEAPI from several aspects including architecture, implementation and security, and address specific challenges in designing compatible interfaces and protecting sensitive data. We describe the functions of SCEAPI including authentication, file transfer and job management for creating, submitting and monitoring, and how to use SCEAPI in an easy-to-use way. Finally, we discuss how to exploit more HPC resources quickly for the ATLAS experiment by implementing the custom ARC compute element based on SCEAPI, and our work shows that SCEAPI is an easy-to-use and effective solution to extend opportunistic HPC resources.

Earth system modelling on system-level heterogeneous architectures: EMAC (version 2.42) on the Dynamical Exascale Entry Platform (DEEP)

NASA Astrophysics Data System (ADS)

Christou, Michalis; Christoudias, Theodoros; Morillo, Julián; Alvarez, Damian; Merx, Hendrik

2016-09-01

We examine an alternative approach to heterogeneous cluster-computing in the many-core era for Earth system models, using the European Centre for Medium-Range Weather Forecasts Hamburg (ECHAM)/Modular Earth Submodel System (MESSy) Atmospheric Chemistry (EMAC) model as a pilot application on the Dynamical Exascale Entry Platform (DEEP). A set of autonomous coprocessors interconnected together, called Booster, complements a conventional HPC Cluster and increases its computing performance, offering extra flexibility to expose multiple levels of parallelism and achieve better scalability. The EMAC model atmospheric chemistry code (Module Efficiently Calculating the Chemistry of the Atmosphere (MECCA)) was taskified with an offload mechanism implemented using OmpSs directives. The model was ported to the MareNostrum 3 supercomputer to allow testing with Intel Xeon Phi accelerators on a production-size machine. The changes proposed in this paper are expected to contribute to the eventual adoption of Cluster-Booster division and Many Integrated Core (MIC) accelerated architectures in presently available implementations of Earth system models, towards exploiting the potential of a fully Exascale-capable platform.

Model-Based Clustering of Regression Time Series Data via APECM -- An AECM Algorithm Sung to an Even Faster Beat

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

Chen, Wei-Chen; Maitra, Ranjan

2011-01-01

We propose a model-based approach for clustering time series regression data in an unsupervised machine learning framework to identify groups under the assumption that each mixture component follows a Gaussian autoregressive regression model of order p. Given the number of groups, the traditional maximum likelihood approach of estimating the parameters using the expectation-maximization (EM) algorithm can be employed, although it is computationally demanding. The somewhat fast tune to the EM folk song provided by the Alternating Expectation Conditional Maximization (AECM) algorithm can alleviate the problem to some extent. In this article, we develop an alternative partial expectation conditional maximization algorithmmore » (APECM) that uses an additional data augmentation storage step to efficiently implement AECM for finite mixture models. Results on our simulation experiments show improved performance in both fewer numbers of iterations and computation time. The methodology is applied to the problem of clustering mutual funds data on the basis of their average annual per cent returns and in the presence of economic indicators.« less

Bond Order Correlations in the 2D Hubbard Model

NASA Astrophysics Data System (ADS)

Moore, Conrad; Abu Asal, Sameer; Yang, Shuxiang; Moreno, Juana; Jarrell, Mark

We use the dynamical cluster approximation to study the bond correlations in the Hubbard model with next nearest neighbor (nnn) hopping to explore the region of the phase diagram where the Fermi liquid phase is separated from the pseudogap phase by the Lifshitz line at zero temperature. We implement the Hirsch-Fye cluster solver that has the advantage of providing direct access to the computation of the bond operators via the decoupling field. In the pseudogap phase, the parallel bond order susceptibility is shown to persist at zero temperature while it vanishes for the Fermi liquid phase which allows the shape of the Lifshitz line to be mapped as a function of filling and nnn hopping. Our cluster solver implements NVIDIA's CUDA language to accelerate the linear algebra of the Quantum Monte Carlo to help alleviate the sign problem by allowing for more Monte Carlo updates to be performed in a reasonable amount of computation time. Work supported by the NSF EPSCoR Cooperative Agreement No. EPS-1003897 with additional support from the Louisiana Board of Regents.

Investigating Phase-Change-Induced Flow in Gas Diffusion Layers in Fuel Cells with X-ray Computed Tomography

DOE PAGES

Shum, Andrew D.; Parkinson, Dilworth Y.; Xiao, Xianghui; ...

2017-10-07

The performance of polymer-electrolyte fuel cells is heavily dependent on proper management of liquid water. One particular reason is that liquid water can collect in the gas diffusion layers (GDLs) blocking the reactant flow to the catalyst layer. This results in increased mass-transport losses. At higher temperatures, evaporation of water becomes a dominant water-removal mechanism and specifically phase-change-induced (PCI) flow is present due to thermal gradients. This study used synchrotron based micro X-ray computed tomography (CT) to visualize and quantify the water distribution within gas diffusion layers subject to a thermal gradient. Plotting saturation as a function of through-plane distancemore » quantitatively shows water redistribution, where water evaporates at hotter locations and condenses in colder locations. The morphology of the 2 GDLs on the micro-scale, as well as evaporating water clusters, are resolved, indicating that the GDL voids are slightly prolate, whereas water clusters are oblate. From the mean radii of water distributions and visual inspection, it is observed that larger water clusters evaporate faster than smaller ones.« less

Predicting Cost/Performance Trade-Offs for Whitney: A Commodity Computing Cluster

NASA Technical Reports Server (NTRS)

Becker, Jeffrey C.; Nitzberg, Bill; VanderWijngaart, Rob F.; Kutler, Paul (Technical Monitor)

1997-01-01

Recent advances in low-end processor and network technology have made it possible to build a "supercomputer" out of commodity components. We develop simple models of the NAS Parallel Benchmarks version 2 (NPB 2) to explore the cost/performance trade-offs involved in building a balanced parallel computer supporting a scientific workload. We develop closed form expressions detailing the number and size of messages sent by each benchmark. Coupling these with measured single processor performance, network latency, and network bandwidth, our models predict benchmark performance to within 30%. A comparison based on total system cost reveals that current commodity technology (200 MHz Pentium Pros with 100baseT Ethernet) is well balanced for the NPBs up to a total system cost of around $1,000,000.

Cluster-specific small airway modeling for imaging-based CFD analysis of pulmonary air flow and particle deposition in COPD smokers

NASA Astrophysics Data System (ADS)

Haghighi, Babak; Choi, Jiwoong; Choi, Sanghun; Hoffman, Eric A.; Lin, Ching-Long

2017-11-01

Accurate modeling of small airway diameters in patients with chronic obstructive pulmonary disease (COPD) is a crucial step toward patient-specific CFD simulations of regional airflow and particle transport. We proposed to use computed tomography (CT) imaging-based cluster membership to identify structural characteristics of airways in each cluster and use them to develop cluster-specific airway diameter models. We analyzed 284 COPD smokers with airflow limitation, and 69 healthy controls. We used multiscale imaging-based cluster analysis (MICA) to classify smokers into 4 clusters. With representative cluster patients and healthy controls, we performed multiple regressions to quantify variation of airway diameters by generation as well as by cluster. The cluster 2 and 4 showed more diameter decrease as generation increases than other clusters. The cluster 4 had more rapid decreases of airway diameters in the upper lobes, while cluster 2 in the lower lobes. We then used these regression models to estimate airway diameters in CT unresolved regions to obtain pressure-volume hysteresis curves using a 1D resistance model. These 1D flow solutions can be used to provide the patient-specific boundary conditions for 3D CFD simulations in COPD patients. Support for this study was provided, in part, by NIH Grants U01-HL114494, R01-HL112986 and S10-RR022421.

A mesh partitioning algorithm for preserving spatial locality in arbitrary geometries

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

Nivarti, Girish V., E-mail: g.nivarti@alumni.ubc.ca; Salehi, M. Mahdi; Bushe, W. Kendal

2015-01-15

Highlights: •An algorithm for partitioning computational meshes is proposed. •The Morton order space-filling curve is modified to achieve improved locality. •A spatial locality metric is defined to compare results with existing approaches. •Results indicate improved performance of the algorithm in complex geometries. -- Abstract: A space-filling curve (SFC) is a proximity preserving linear mapping of any multi-dimensional space and is widely used as a clustering tool. Equi-sized partitioning of an SFC ignores the loss in clustering quality that occurs due to inaccuracies in the mapping. Often, this results in poor locality within partitions, especially for the conceptually simple, Morton ordermore » curves. We present a heuristic that improves partition locality in arbitrary geometries by slicing a Morton order curve at points where spatial locality is sacrificed. In addition, we develop algorithms that evenly distribute points to the extent possible while maintaining spatial locality. A metric is defined to estimate relative inter-partition contact as an indicator of communication in parallel computing architectures. Domain partitioning tests have been conducted on geometries relevant to turbulent reactive flow simulations. The results obtained highlight the performance of our method as an unsupervised and computationally inexpensive domain partitioning tool.« less

Development of self-compressing BLSOM for comprehensive analysis of big sequence data.

PubMed

Kikuchi, Akihito; Ikemura, Toshimichi; Abe, Takashi

2015-01-01

With the remarkable increase in genomic sequence data from various organisms, novel tools are needed for comprehensive analyses of available big sequence data. We previously developed a Batch-Learning Self-Organizing Map (BLSOM), which can cluster genomic fragment sequences according to phylotype solely dependent on oligonucleotide composition and applied to genome and metagenomic studies. BLSOM is suitable for high-performance parallel-computing and can analyze big data simultaneously, but a large-scale BLSOM needs a large computational resource. We have developed Self-Compressing BLSOM (SC-BLSOM) for reduction of computation time, which allows us to carry out comprehensive analysis of big sequence data without the use of high-performance supercomputers. The strategy of SC-BLSOM is to hierarchically construct BLSOMs according to data class, such as phylotype. The first-layer BLSOM was constructed with each of the divided input data pieces that represents the data subclass, such as phylotype division, resulting in compression of the number of data pieces. The second BLSOM was constructed with a total of weight vectors obtained in the first-layer BLSOMs. We compared SC-BLSOM with the conventional BLSOM by analyzing bacterial genome sequences. SC-BLSOM could be constructed faster than BLSOM and cluster the sequences according to phylotype with high accuracy, showing the method's suitability for efficient knowledge discovery from big sequence data.

Initial experience with computer aided detection for microcalcification in digital breast tomosynthesis

NASA Astrophysics Data System (ADS)

Harkness, E. F.; Lim, Y. Y.; Wilson, M. W.; Haq, R.; Zhou, J.; Tate, C.; Maxwell, A. J.; Astley, S. M.; Gilbert, F. J.

2015-03-01

Digital breast tomosynthesis (DBT) addresses limitations of 2-D projection imaging for detection of masses. Microcalcification clusters may be more difficult to appreciate in DBT as individual calcifications within clusters may appear on different slices. This research aims to evaluate the performance of ImageChecker 3D Calc CAD v1.0. Women were recruited as part of the TOMMY trial. From the trial, 169 were included in this study. The DBT images were processed with the computer aided detection (CAD) algorithm. Three consultant radiologists reviewed the images and recorded whether CAD prompts were on or off target. 79/80 (98.8%) malignant cases had a prompt on the area of microcalcification. In these cases, there were 1-15 marks (median 5) with the majority of false prompts (n=326/431) due to benign (68%) and vascular (24%) calcifications. Of 89 normal/benign cases, there were 1-13 prompts (median 3), 27 (30%) had no prompts and the majority of false prompts (n=238) were benign (77%) calcifications. CAD is effective in prompting malignant microcalcification clusters and may overcome the difficulty of detecting clusters in slice images. Although there was a high rate of false prompts, further advances in the software may improve specificity.

HPC in a HEP lab: lessons learned from setting up cost-effective HPC clusters

NASA Astrophysics Data System (ADS)

Husejko, Michal; Agtzidis, Ioannis; Baehler, Pierre; Dul, Tadeusz; Evans, John; Himyr, Nils; Meinhard, Helge

2015-12-01

In this paper we present our findings gathered during the evaluation and testing of Windows Server High-Performance Computing (Windows HPC) in view of potentially using it as a production HPC system for engineering applications. The Windows HPC package, an extension of Microsofts Windows Server product, provides all essential interfaces, utilities and management functionality for creating, operating and monitoring a Windows-based HPC cluster infrastructure. The evaluation and test phase was focused on verifying the functionalities of Windows HPC, its performance, support of commercial tools and the integration with the users work environment. We describe constraints imposed by the way the CERN Data Centre is operated, licensing for engineering tools and scalability and behaviour of the HPC engineering applications used at CERN. We will present an initial set of requirements, which were created based on the above constraints and requests from the CERN engineering user community. We will explain how we have configured Windows HPC clusters to provide job scheduling functionalities required to support the CERN engineering user community, quality of service, user- and project-based priorities, and fair access to limited resources. Finally, we will present several performance tests we carried out to verify Windows HPC performance and scalability.

Structure and stability of clusters of β-alanine in the gas phase: importance of the nature of intermolecular interactions.

PubMed

Piekarski, Dariusz Grzegorz; Díaz-Tendero, Sergio

2017-02-15

We present a theoretical study of neutral clusters of β-alanine molecules in the gas phase, (β-ala) n n ≤ 5. Classical molecular dynamics simulations carried out with different internal excitation energies provide information on the clusters formation and their thermal decomposition limits. We also present an assessment study performed with different families of density functionals using the dimer, (β-ala) 2 , as a benchmark system. The M06-2X functional provides the best agreement in geometries and relative energies in comparison with the reference values computed with the MP2 and CCSD(T) methods. The structure, stability, dissociation energies and vertical ionization potentials of the studied clusters have been investigated using this functional in combination with the 6-311++G(d,p) basis set. An exhaustive analysis of intermolecular interactions is also presented. These results provide new insights into the stability, interaction nature and formation mechanisms of clusters of amino acids in the gas phase.

Automatic categorization of anatomical landmark-local appearances based on diffeomorphic demons and spectral clustering for constructing detector ensembles.

PubMed

Hanaoka, Shouhei; Masutani, Yoshitaka; Nemoto, Mitsutaka; Nomura, Yukihiro; Yoshikawa, Takeharu; Hayashi, Naoto; Ohtomo, Kuni

2012-01-01

A method for categorizing landmark-local appearances extracted from computed tomography (CT) datasets is presented. Anatomical landmarks in the human body inevitably have inter-individual variations that cause difficulty in automatic landmark detection processes. The goal of this study is to categorize subjects (i.e., training datasets) according to local shape variations of such a landmark so that each subgroup has less shape variation and thus the machine learning of each landmark detector is much easier. The similarity between each subject pair is measured based on the non-rigid registration result between them. These similarities are used by the spectral clustering process. After the clustering, all training datasets in each cluster, as well as synthesized intermediate images calculated from all subject-pairs in the cluster, are used to train the corresponding subgroup detector. All of these trained detectors compose a detector ensemble to detect the target landmark. Evaluation with clinical CT datasets showed great improvement in the detection performance.

Coresets vs clustering: comparison of methods for redundancy reduction in very large white matter fiber sets

NASA Astrophysics Data System (ADS)

Alexandroni, Guy; Zimmerman Moreno, Gali; Sochen, Nir; Greenspan, Hayit

2016-03-01

Recent advances in Diffusion Weighted Magnetic Resonance Imaging (DW-MRI) of white matter in conjunction with improved tractography produce impressive reconstructions of White Matter (WM) pathways. These pathways (fiber sets) often contain hundreds of thousands of fibers, or more. In order to make fiber based analysis more practical, the fiber set needs to be preprocessed to eliminate redundancies and to keep only essential representative fibers. In this paper we demonstrate and compare two distinctive frameworks for selecting this reduced set of fibers. The first framework entails pre-clustering the fibers using k-means, followed by Hierarchical Clustering and replacing each cluster with one representative. For the second clustering stage seven distance metrics were evaluated. The second framework is based on an efficient geometric approximation paradigm named coresets. Coresets present a new approach to optimization and have huge success especially in tasks requiring large computation time and/or memory. We propose a modified version of the coresets algorithm, Density Coreset. It is used for extracting the main fibers from dense datasets, leaving a small set that represents the main structures and connectivity of the brain. A novel approach, based on a 3D indicator structure, is used for comparing the frameworks. This comparison was applied to High Angular Resolution Diffusion Imaging (HARDI) scans of 4 healthy individuals. We show that among the clustering based methods, that cosine distance gives the best performance. In comparing the clustering schemes with coresets, Density Coreset method achieves the best performance.

Structure-sequence based analysis for identification of conserved regions in proteins

DOEpatents

Zemla, Adam T; Zhou, Carol E; Lam, Marisa W; Smith, Jason R; Pardes, Elizabeth

2013-05-28

Disclosed are computational methods, and associated hardware and software products for scoring conservation in a protein structure based on a computationally identified family or cluster of protein structures. A method of computationally identifying a family or cluster of protein structures in also disclosed herein.

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.

High Performance Semantic Factoring of Giga-Scale Semantic Graph Databases

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

Joslyn, Cliff A.; Adolf, Robert D.; Al-Saffar, Sinan

2010-10-04

As semantic graph database technology grows to address components ranging from extant large triple stores to SPARQL endpoints over SQL-structured relational databases, it will become increasingly important to be able to bring high performance computational resources to bear on their analysis, interpretation, and visualization, especially with respect to their innate semantic structure. Our research group built a novel high performance hybrid system comprising computational capability for semantic graph database processing utilizing the large multithreaded architecture of the Cray XMT platform, conventional clusters, and large data stores. In this paper we describe that architecture, and present the results of our deployingmore » that for the analysis of the Billion Triple dataset with respect to its semantic factors.« less

Accelerating DNA analysis applications on GPU clusters

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

Tumeo, Antonino; Villa, Oreste

DNA analysis is an emerging application of high performance bioinformatic. Modern sequencing machinery are able to provide, in few hours, large input streams of data which needs to be matched against exponentially growing databases known fragments. The ability to recognize these patterns effectively and fastly may allow extending the scale and the reach of the investigations performed by biology scientists. Aho-Corasick is an exact, multiple pattern matching algorithm often at the base of this application. High performance systems are a promising platform to accelerate this algorithm, which is computationally intensive but also inherently parallel. Nowadays, high performance systems also includemore » heterogeneous processing elements, such as Graphic Processing Units (GPUs), to further accelerate parallel algorithms. Unfortunately, the Aho-Corasick algorithm exhibits large performance variabilities, depending on the size of the input streams, on the number of patterns to search and on the number of matches, and poses significant challenges on current high performance software and hardware implementations. An adequate mapping of the algorithm on the target architecture, coping with the limit of the underlining hardware, is required to reach the desired high throughputs. Load balancing also plays a crucial role when considering the limited bandwidth among the nodes of these systems. In this paper we present an efficient implementation of the Aho-Corasick algorithm for high performance clusters accelerated with GPUs. We discuss how we partitioned and adapted the algorithm to fit the Tesla C1060 GPU and then present a MPI based implementation for a heterogeneous high performance cluster. We compare this implementation to MPI and MPI with pthreads based implementations for a homogeneous cluster of x86 processors, discussing the stability vs. the performance and the scaling of the solutions, taking into consideration aspects such as the bandwidth among the different nodes.« less

Fast distributed large-pixel-count hologram computation using a GPU cluster.

PubMed

Pan, Yuechao; Xu, Xuewu; Liang, Xinan

2013-09-10

Large-pixel-count holograms are one essential part for big size holographic three-dimensional (3D) display, but the generation of such holograms is computationally demanding. In order to address this issue, we have built a graphics processing unit (GPU) cluster with 32.5 Tflop/s computing power and implemented distributed hologram computation on it with speed improvement techniques, such as shared memory on GPU, GPU level adaptive load balancing, and node level load distribution. Using these speed improvement techniques on the GPU cluster, we have achieved 71.4 times computation speed increase for 186M-pixel holograms. Furthermore, we have used the approaches of diffraction limits and subdivision of holograms to overcome the GPU memory limit in computing large-pixel-count holograms. 745M-pixel and 1.80G-pixel holograms were computed in 343 and 3326 s, respectively, for more than 2 million object points with RGB colors. Color 3D objects with 1.02M points were successfully reconstructed from 186M-pixel hologram computed in 8.82 s with all the above three speed improvement techniques. It is shown that distributed hologram computation using a GPU cluster is a promising approach to increase the computation speed of large-pixel-count holograms for large size holographic display.

A New Soft Computing Method for K-Harmonic Means Clustering.

PubMed

Yeh, Wei-Chang; Jiang, Yunzhi; Chen, Yee-Fen; Chen, Zhe

2016-01-01

The K-harmonic means clustering algorithm (KHM) is a new clustering method used to group data such that the sum of the harmonic averages of the distances between each entity and all cluster centroids is minimized. Because it is less sensitive to initialization than K-means (KM), many researchers have recently been attracted to studying KHM. In this study, the proposed iSSO-KHM is based on an improved simplified swarm optimization (iSSO) and integrates a variable neighborhood search (VNS) for KHM clustering. As evidence of the utility of the proposed iSSO-KHM, we present extensive computational results on eight benchmark problems. From the computational results, the comparison appears to support the superiority of the proposed iSSO-KHM over previously developed algorithms for all experiments in the literature.

CLASSIFY: A Group Teaching Exercise in Microbial Identification and Numerical Taxonomy Using a Commodore 64 Microcomputer.

ERIC Educational Resources Information Center

Soddell, J. A.; Seviour, R. J.

1985-01-01

Describes an exercise which uses a computer program (written for Commodore 64 microcomputers) that accepts data obtained from identifying bacteria, calculates similarity coefficients, and performs single linkage cluster analysis. Includes a program for simulating bacterial cultures for students who should not handle pathogenic microorganisms. (JN)

Inputs to Vocational-Technical Education from Occupational Research.

ERIC Educational Resources Information Center

Christal, Raymond E.

The Air Force has developed a data bank which classifies job descriptions into occupational clusters. A retrieval system, also recently developed computes and publishes a consolidated description of the work being performed by any group which can be defined in terms of the background information. Other retrieval programs are available which enable…

Health and performance monitoring of the online computer cluster of CMS

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

Bauer, G.; et al.

2012-01-01

The CMS experiment at the LHC features over 2'500 devices that need constant monitoring in order to ensure proper data taking. The monitoring solution has been migrated from Nagios to Icinga, with several useful plugins. The motivations behind the migration and the selection of the plugins are discussed.

Deconvoluting simulated metagenomes: the performance of hard- and soft- clustering algorithms applied to metagenomic chromosome conformation capture (3C)

PubMed Central

DeMaere, Matthew Z.

2016-01-01

Background Chromosome conformation capture, coupled with high throughput DNA sequencing in protocols like Hi-C and 3C-seq, has been proposed as a viable means of generating data to resolve the genomes of microorganisms living in naturally occuring environments. Metagenomic Hi-C and 3C-seq datasets have begun to emerge, but the feasibility of resolving genomes when closely related organisms (strain-level diversity) are present in the sample has not yet been systematically characterised. Methods We developed a computational simulation pipeline for metagenomic 3C and Hi-C sequencing to evaluate the accuracy of genomic reconstructions at, above, and below an operationally defined species boundary. We simulated datasets and measured accuracy over a wide range of parameters. Five clustering algorithms were evaluated (2 hard, 3 soft) using an adaptation of the extended B-cubed validation measure. Results When all genomes in a sample are below 95% sequence identity, all of the tested clustering algorithms performed well. When sequence data contains genomes above 95% identity (our operational definition of strain-level diversity), a naive soft-clustering extension of the Louvain method achieves the highest performance. Discussion Previously, only hard-clustering algorithms have been applied to metagenomic 3C and Hi-C data, yet none of these perform well when strain-level diversity exists in a metagenomic sample. Our simple extension of the Louvain method performed the best in these scenarios, however, accuracy remained well below the levels observed for samples without strain-level diversity. Strain resolution is also highly dependent on the amount of available 3C sequence data, suggesting that depth of sequencing must be carefully considered during experimental design. Finally, there appears to be great scope to improve the accuracy of strain resolution through further algorithm development. PMID:27843713

Multi-class ERP-based BCI data analysis using a discriminant space self-organizing map.

PubMed

Onishi, Akinari; Natsume, Kiyohisa

2014-01-01

Emotional or non-emotional image stimulus is recently applied to event-related potential (ERP) based brain computer interfaces (BCI). Though the classification performance is over 80% in a single trial, a discrimination between those ERPs has not been considered. In this research we tried to clarify the discriminability of four-class ERP-based BCI target data elicited by desk, seal, spider images and letter intensifications. A conventional self organizing map (SOM) and newly proposed discriminant space SOM (ds-SOM) were applied, then the discriminabilites were visualized. We also classify all pairs of those ERPs by stepwise linear discriminant analysis (SWLDA) and verify the visualization of discriminabilities. As a result, the ds-SOM showed understandable visualization of the data with a shorter computational time than the traditional SOM. We also confirmed the clear boundary between the letter cluster and the other clusters. The result was coherent with the classification performances by SWLDA. The method might be helpful not only for developing a new BCI paradigm, but also for the big data analysis.

Parallel Numerical Simulations of Water Reservoirs

NASA Astrophysics Data System (ADS)

Torres, Pedro; Mangiavacchi, Norberto

2010-11-01

The study of the water flow and scalar transport in water reservoirs is important for the determination of the water quality during the initial stages of the reservoir filling and during the life of the reservoir. For this scope, a parallel 2D finite element code for solving the incompressible Navier-Stokes equations coupled with scalar transport was implemented using the message-passing programming model, in order to perform simulations of hidropower water reservoirs in a computer cluster environment. The spatial discretization is based on the MINI element that satisfies the Babuska-Brezzi (BB) condition, which provides sufficient conditions for a stable mixed formulation. All the distributed data structures needed in the different stages of the code, such as preprocessing, solving and post processing, were implemented using the PETSc library. The resulting linear systems for the velocity and the pressure fields were solved using the projection method, implemented by an approximate block LU factorization. In order to increase the parallel performance in the solution of the linear systems, we employ the static condensation method for solving the intermediate velocity at vertex and centroid nodes separately. We compare performance results of the static condensation method with the approach of solving the complete system. In our tests the static condensation method shows better performance for large problems, at the cost of an increased memory usage. Performance results for other intensive parts of the code in a computer cluster are also presented.

Hierarchical trie packet classification algorithm based on expectation-maximization clustering.

PubMed

Bi, Xia-An; Zhao, Junxia

2017-01-01

With the development of computer network bandwidth, packet classification algorithms which are able to deal with large-scale rule sets are in urgent need. Among the existing algorithms, researches on packet classification algorithms based on hierarchical trie have become an important packet classification research branch because of their widely practical use. Although hierarchical trie is beneficial to save large storage space, it has several shortcomings such as the existence of backtracking and empty nodes. This paper proposes a new packet classification algorithm, Hierarchical Trie Algorithm Based on Expectation-Maximization Clustering (HTEMC). Firstly, this paper uses the formalization method to deal with the packet classification problem by means of mapping the rules and data packets into a two-dimensional space. Secondly, this paper uses expectation-maximization algorithm to cluster the rules based on their aggregate characteristics, and thereby diversified clusters are formed. Thirdly, this paper proposes a hierarchical trie based on the results of expectation-maximization clustering. Finally, this paper respectively conducts simulation experiments and real-environment experiments to compare the performances of our algorithm with other typical algorithms, and analyzes the results of the experiments. The hierarchical trie structure in our algorithm not only adopts trie path compression to eliminate backtracking, but also solves the problem of low efficiency of trie updates, which greatly improves the performance of the algorithm.

Hydrogen bonding in water clusters and their ionized counterparts.

PubMed

Neela, Y Indra; Mahadevi, A Subha; Sastry, G Narahari

2010-12-30

Ab initio and DFT computations were carried out on four distinct hydrogen-bonded arrangements of water clusters (H(2)O)(n), n = 2-20, represented as W1D, W2D, W2DH, and W3D. The variation in the strength of hydrogen bond as a function of the chain length is studied. In all the four cases, there is a substantial cooperative interaction, albeit in different degrees. The effect of basis set superposition error (BSSE) on the complexation energy of water clusters has been analyzed. Atoms in molecules (AIM) analysis performed to evaluate the nature of the hydrogen bonding shows a high correlation between hydrogen bond strength and the trends in complexation energy. Solvated water clusters exhibit lower complexation energies compared to corresponding gas-phase geometries on PCM (polarized continuum model) optimization. The feasibility of stripping an electron or addition of an electron increases dramatically as the cluster size increases. Although W3D caged structures are stable for neutral clusters, the helical W2DH arrangement appeared to be an optimal choice for its ionized counterparts.

Machine learning approaches for estimation of prediction interval for the model output.

PubMed

Shrestha, Durga L; Solomatine, Dimitri P

2006-03-01

A novel method for estimating prediction uncertainty using machine learning techniques is presented. Uncertainty is expressed in the form of the two quantiles (constituting the prediction interval) of the underlying distribution of prediction errors. The idea is to partition the input space into different zones or clusters having similar model errors using fuzzy c-means clustering. The prediction interval is constructed for each cluster on the basis of empirical distributions of the errors associated with all instances belonging to the cluster under consideration and propagated from each cluster to the examples according to their membership grades in each cluster. Then a regression model is built for in-sample data using computed prediction limits as targets, and finally, this model is applied to estimate the prediction intervals (limits) for out-of-sample data. The method was tested on artificial and real hydrologic data sets using various machine learning techniques. Preliminary results show that the method is superior to other methods estimating the prediction interval. A new method for evaluating performance for estimating prediction interval is proposed as well.

Massively parallel quantum computer simulator

NASA Astrophysics Data System (ADS)

De Raedt, K.; Michielsen, K.; De Raedt, H.; Trieu, B.; Arnold, G.; Richter, M.; Lippert, Th.; Watanabe, H.; Ito, N.

2007-01-01

We describe portable software to simulate universal quantum computers on massive parallel computers. We illustrate the use of the simulation software by running various quantum algorithms on different computer architectures, such as a IBM BlueGene/L, a IBM Regatta p690+, a Hitachi SR11000/J1, a Cray X1E, a SGI Altix 3700 and clusters of PCs running Windows XP. We study the performance of the software by simulating quantum computers containing up to 36 qubits, using up to 4096 processors and up to 1 TB of memory. Our results demonstrate that the simulator exhibits nearly ideal scaling as a function of the number of processors and suggest that the simulation software described in this paper may also serve as benchmark for testing high-end parallel computers.

Mixing HTC and HPC Workloads with HTCondor and Slurm

NASA Astrophysics Data System (ADS)

Hollowell, C.; Barnett, J.; Caramarcu, C.; Strecker-Kellogg, W.; Wong, A.; Zaytsev, A.

2017-10-01

Traditionally, the RHIC/ATLAS Computing Facility (RACF) at Brookhaven National Laboratory (BNL) has only maintained High Throughput Computing (HTC) resources for our HEP/NP user community. We’ve been using HTCondor as our batch system for many years, as this software is particularly well suited for managing HTC processor farm resources. Recently, the RACF has also begun to design/administrate some High Performance Computing (HPC) systems for a multidisciplinary user community at BNL. In this paper, we’ll discuss our experiences using HTCondor and Slurm in an HPC context, and our facility’s attempts to allow our HTC and HPC processing farms/clusters to make opportunistic use of each other’s computing resources.

Design and implementation of a reliable and cost-effective cloud computing infrastructure: the INFN Napoli experience

NASA Astrophysics Data System (ADS)

Capone, V.; Esposito, R.; Pardi, S.; Taurino, F.; Tortone, G.

2012-12-01

Over the last few years we have seen an increasing number of services and applications needed to manage and maintain cloud computing facilities. This is particularly true for computing in high energy physics, which often requires complex configurations and distributed infrastructures. In this scenario a cost effective rationalization and consolidation strategy is the key to success in terms of scalability and reliability. In this work we describe an IaaS (Infrastructure as a Service) cloud computing system, with high availability and redundancy features, which is currently in production at INFN-Naples and ATLAS Tier-2 data centre. The main goal we intended to achieve was a simplified method to manage our computing resources and deliver reliable user services, reusing existing hardware without incurring heavy costs. A combined usage of virtualization and clustering technologies allowed us to consolidate our services on a small number of physical machines, reducing electric power costs. As a result of our efforts we developed a complete solution for data and computing centres that can be easily replicated using commodity hardware. Our architecture consists of 2 main subsystems: a clustered storage solution, built on top of disk servers running GlusterFS file system, and a virtual machines execution environment. GlusterFS is a network file system able to perform parallel writes on multiple disk servers, providing this way live replication of data. High availability is also achieved via a network configuration using redundant switches and multiple paths between hypervisor hosts and disk servers. We also developed a set of management scripts to easily perform basic system administration tasks such as automatic deployment of new virtual machines, adaptive scheduling of virtual machines on hypervisor hosts, live migration and automated restart in case of hypervisor failures.

The Computation of Orthogonal Independent Cluster Solutions and Their Oblique Analogs in Factor Analysis.

ERIC Educational Resources Information Center

Hofmann, Richard J.

A very general model for the computation of independent cluster solutions in factor analysis is presented. The model is discussed as being either orthogonal or oblique. Furthermore, it is demonstrated that for every orthogonal independent cluster solution there is an oblique analog. Using three illustrative examples, certain generalities are made…

Accelerating the coupled-cluster singles and doubles method using the chain-of-sphere approximation

NASA Astrophysics Data System (ADS)

Dutta, Achintya Kumar; Neese, Frank; Izsák, Róbert

2018-06-01

In this paper, we present a chain-of-sphere implementation of the external exchange term, the computational bottleneck of coupled-cluster calculations at the singles and doubles level. This implementation is compared to standard molecular orbital, atomic orbital and resolution of identity implementations of the same term within the ORCA package and turns out to be the most efficient one for larger molecules, with a better accuracy than the resolution-of-identity approximation. Furthermore, it becomes possible to perform a canonical CC calculation on a tetramer of nucleobases in 17 days, 20 hours.

Image Registration Algorithm Based on Parallax Constraint and Clustering Analysis

NASA Astrophysics Data System (ADS)

Wang, Zhe; Dong, Min; Mu, Xiaomin; Wang, Song

2018-01-01

To resolve the problem of slow computation speed and low matching accuracy in image registration, a new image registration algorithm based on parallax constraint and clustering analysis is proposed. Firstly, Harris corner detection algorithm is used to extract the feature points of two images. Secondly, use Normalized Cross Correlation (NCC) function to perform the approximate matching of feature points, and the initial feature pair is obtained. Then, according to the parallax constraint condition, the initial feature pair is preprocessed by K-means clustering algorithm, which is used to remove the feature point pairs with obvious errors in the approximate matching process. Finally, adopt Random Sample Consensus (RANSAC) algorithm to optimize the feature points to obtain the final feature point matching result, and the fast and accurate image registration is realized. The experimental results show that the image registration algorithm proposed in this paper can improve the accuracy of the image matching while ensuring the real-time performance of the algorithm.

Generalization of Clustering Coefficients to Signed Correlation Networks

PubMed Central

Costantini, Giulio; Perugini, Marco

2014-01-01

The recent interest in network analysis applications in personality psychology and psychopathology has put forward new methodological challenges. Personality and psychopathology networks are typically based on correlation matrices and therefore include both positive and negative edge signs. However, some applications of network analysis disregard negative edges, such as computing clustering coefficients. In this contribution, we illustrate the importance of the distinction between positive and negative edges in networks based on correlation matrices. The clustering coefficient is generalized to signed correlation networks: three new indices are introduced that take edge signs into account, each derived from an existing and widely used formula. The performances of the new indices are illustrated and compared with the performances of the unsigned indices, both on a signed simulated network and on a signed network based on actual personality psychology data. The results show that the new indices are more resistant to sample variations in correlation networks and therefore have higher convergence compared with the unsigned indices both in simulated networks and with real data. PMID:24586367

The [(AI 2O 3) 2] - Anion Cluster: Electron Localization-Delocalization Isomerism

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

Sierka, Marek; Dobler, Jens; Sauer, Joachim

2009-10-05

Three-dimensional bulk alumina and its two-dimensional thin films show great structural diversity, posing considerable challenges to their experimental structural characterization and computational modeling. Recently, structural diversity has also been demonstrated for zerodimensional gas phase aluminum oxide clusters. Mass-selected clusters not only make systematic studies of the structural and electronic properties as a function of size possible, but lately have also emerged as powerful molecular models of complex surfaces and solid catalysts. In particular, the [(Al 2O 3) 3-5] + clusters were the first example of polynuclear maingroup metal oxide cluster that are able to thermally activate CH 4. Over themore » past decades gas phase aluminum oxide clusters have been extensively studied both experimentally and computationally, but definitive structural assignments were made for only a handful of them: the planar [Al 3O 3] - and [Al 5O 4] - cluster anions, and the [(Al 2O 3) 1-4(AlO)] + cluster cations. For stoichiometric clusters only the atomic structures of [(Al 2O 3) 4] +/0 have been nambiguously resolved. Here we report on the structures of the [(Al 2O 3) 2] -/0 clusters combining photoelectron spectroscopy (PES) and quantum chemical calculations employing a genetic algorithm as a global optimization technique. The [(Al 2O 3) 2] - cluster anion show energetically close lying but structurally distinct cage and sheet-like isomers which differ by delocalization/localization of the extra electron. The experimental results are crucial for benchmarking the different computational methods applied with respect to a proper description of electron localization and the relative energies for the isomers which will be of considerable value for future computational studies of aluminum oxide and related systems.« less

A data colocation grid framework for big data medical image processing: backend design

NASA Astrophysics Data System (ADS)

Bao, Shunxing; Huo, Yuankai; Parvathaneni, Prasanna; Plassard, Andrew J.; Bermudez, Camilo; Yao, Yuang; Lyu, Ilwoo; Gokhale, Aniruddha; Landman, Bennett A.

2018-03-01

When processing large medical imaging studies, adopting high performance grid computing resources rapidly becomes important. We recently presented a "medical image processing-as-a-service" grid framework that offers promise in utilizing the Apache Hadoop ecosystem and HBase for data colocation by moving computation close to medical image storage. However, the framework has not yet proven to be easy to use in a heterogeneous hardware environment. Furthermore, the system has not yet validated when considering variety of multi-level analysis in medical imaging. Our target design criteria are (1) improving the framework's performance in a heterogeneous cluster, (2) performing population based summary statistics on large datasets, and (3) introducing a table design scheme for rapid NoSQL query. In this paper, we present a heuristic backend interface application program interface (API) design for Hadoop and HBase for Medical Image Processing (HadoopBase-MIP). The API includes: Upload, Retrieve, Remove, Load balancer (for heterogeneous cluster) and MapReduce templates. A dataset summary statistic model is discussed and implemented by MapReduce paradigm. We introduce a HBase table scheme for fast data query to better utilize the MapReduce model. Briefly, 5153 T1 images were retrieved from a university secure, shared web database and used to empirically access an in-house grid with 224 heterogeneous CPU cores. Three empirical experiments results are presented and discussed: (1) load balancer wall-time improvement of 1.5-fold compared with a framework with built-in data allocation strategy, (2) a summary statistic model is empirically verified on grid framework and is compared with the cluster when deployed with a standard Sun Grid Engine (SGE), which reduces 8-fold of wall clock time and 14-fold of resource time, and (3) the proposed HBase table scheme improves MapReduce computation with 7 fold reduction of wall time compare with a naïve scheme when datasets are relative small. The source code and interfaces have been made publicly available.

A Data Colocation Grid Framework for Big Data Medical Image Processing: Backend Design.

PubMed

Bao, Shunxing; Huo, Yuankai; Parvathaneni, Prasanna; Plassard, Andrew J; Bermudez, Camilo; Yao, Yuang; Lyu, Ilwoo; Gokhale, Aniruddha; Landman, Bennett A

2018-03-01

When processing large medical imaging studies, adopting high performance grid computing resources rapidly becomes important. We recently presented a "medical image processing-as-a-service" grid framework that offers promise in utilizing the Apache Hadoop ecosystem and HBase for data colocation by moving computation close to medical image storage. However, the framework has not yet proven to be easy to use in a heterogeneous hardware environment. Furthermore, the system has not yet validated when considering variety of multi-level analysis in medical imaging. Our target design criteria are (1) improving the framework's performance in a heterogeneous cluster, (2) performing population based summary statistics on large datasets, and (3) introducing a table design scheme for rapid NoSQL query. In this paper, we present a heuristic backend interface application program interface (API) design for Hadoop & HBase for Medical Image Processing (HadoopBase-MIP). The API includes: Upload, Retrieve, Remove, Load balancer (for heterogeneous cluster) and MapReduce templates. A dataset summary statistic model is discussed and implemented by MapReduce paradigm. We introduce a HBase table scheme for fast data query to better utilize the MapReduce model. Briefly, 5153 T1 images were retrieved from a university secure, shared web database and used to empirically access an in-house grid with 224 heterogeneous CPU cores. Three empirical experiments results are presented and discussed: (1) load balancer wall-time improvement of 1.5-fold compared with a framework with built-in data allocation strategy, (2) a summary statistic model is empirically verified on grid framework and is compared with the cluster when deployed with a standard Sun Grid Engine (SGE), which reduces 8-fold of wall clock time and 14-fold of resource time, and (3) the proposed HBase table scheme improves MapReduce computation with 7 fold reduction of wall time compare with a naïve scheme when datasets are relative small. The source code and interfaces have been made publicly available.

A Data Colocation Grid Framework for Big Data Medical Image Processing: Backend Design

PubMed Central

Huo, Yuankai; Parvathaneni, Prasanna; Plassard, Andrew J.; Bermudez, Camilo; Yao, Yuang; Lyu, Ilwoo; Gokhale, Aniruddha; Landman, Bennett A.

2018-01-01

When processing large medical imaging studies, adopting high performance grid computing resources rapidly becomes important. We recently presented a "medical image processing-as-a-service" grid framework that offers promise in utilizing the Apache Hadoop ecosystem and HBase for data colocation by moving computation close to medical image storage. However, the framework has not yet proven to be easy to use in a heterogeneous hardware environment. Furthermore, the system has not yet validated when considering variety of multi-level analysis in medical imaging. Our target design criteria are (1) improving the framework’s performance in a heterogeneous cluster, (2) performing population based summary statistics on large datasets, and (3) introducing a table design scheme for rapid NoSQL query. In this paper, we present a heuristic backend interface application program interface (API) design for Hadoop & HBase for Medical Image Processing (HadoopBase-MIP). The API includes: Upload, Retrieve, Remove, Load balancer (for heterogeneous cluster) and MapReduce templates. A dataset summary statistic model is discussed and implemented by MapReduce paradigm. We introduce a HBase table scheme for fast data query to better utilize the MapReduce model. Briefly, 5153 T1 images were retrieved from a university secure, shared web database and used to empirically access an in-house grid with 224 heterogeneous CPU cores. Three empirical experiments results are presented and discussed: (1) load balancer wall-time improvement of 1.5-fold compared with a framework with built-in data allocation strategy, (2) a summary statistic model is empirically verified on grid framework and is compared with the cluster when deployed with a standard Sun Grid Engine (SGE), which reduces 8-fold of wall clock time and 14-fold of resource time, and (3) the proposed HBase table scheme improves MapReduce computation with 7 fold reduction of wall time compare with a naïve scheme when datasets are relative small. The source code and interfaces have been made publicly available. PMID:29887668

Characterization of computer network events through simultaneous feature selection and clustering of intrusion alerts

NASA Astrophysics Data System (ADS)

Chen, Siyue; Leung, Henry; Dondo, Maxwell

2014-05-01

As computer network security threats increase, many organizations implement multiple Network Intrusion Detection Systems (NIDS) to maximize the likelihood of intrusion detection and provide a comprehensive understanding of intrusion activities. However, NIDS trigger a massive number of alerts on a daily basis. This can be overwhelming for computer network security analysts since it is a slow and tedious process to manually analyse each alert produced. Thus, automated and intelligent clustering of alerts is important to reveal the structural correlation of events by grouping alerts with common features. As the nature of computer network attacks, and therefore alerts, is not known in advance, unsupervised alert clustering is a promising approach to achieve this goal. We propose a joint optimization technique for feature selection and clustering to aggregate similar alerts and to reduce the number of alerts that analysts have to handle individually. More precisely, each identified feature is assigned a binary value, which reflects the feature's saliency. This value is treated as a hidden variable and incorporated into a likelihood function for clustering. Since computing the optimal solution of the likelihood function directly is analytically intractable, we use the Expectation-Maximisation (EM) algorithm to iteratively update the hidden variable and use it to maximize the expected likelihood. Our empirical results, using a labelled Defense Advanced Research Projects Agency (DARPA) 2000 reference dataset, show that the proposed method gives better results than the EM clustering without feature selection in terms of the clustering accuracy.

Hybrid MPI+OpenMP Programming of an Overset CFD Solver and Performance Investigations

NASA Technical Reports Server (NTRS)

Djomehri, M. Jahed; Jin, Haoqiang H.; Biegel, Bryan (Technical Monitor)

2002-01-01

This report describes a two level parallelization of a Computational Fluid Dynamic (CFD) solver with multi-zone overset structured grids. The approach is based on a hybrid MPI+OpenMP programming model suitable for shared memory and clusters of shared memory machines. The performance investigations of the hybrid application on an SGI Origin2000 (O2K) machine is reported using medium and large scale test problems.

Detecting Genomic Clustering of Risk Variants from Sequence Data: Cases vs. Controls

PubMed Central

Schaid, Daniel J.; Sinnwell, Jason P.; McDonnell, Shannon K.; Thibodeau, Stephen N.

2013-01-01

As the ability to measure dense genetic markers approaches the limit of the DNA sequence itself, taking advantage of possible clustering of genetic variants in, and around, a gene would benefit genetic association analyses, and likely provide biological insights. The greatest benefit might be realized when multiple rare variants cluster in a functional region. Several statistical tests have been developed, one of which is based on the popular Kulldorff scan statistic for spatial clustering of disease. We extended another popular spatial clustering method – Tango’s statistic – to genomic sequence data. An advantage of Tango’s method is that it is rapid to compute, and when single test statistic is computed, its distribution is well approximated by a scaled chi-square distribution, making computation of p-values very rapid. We compared the Type-I error rates and power of several clustering statistics, as well as the omnibus sequence kernel association test (SKAT). Although our version of Tango’s statistic, which we call “Kernel Distance” statistic, took approximately half the time to compute than the Kulldorff scan statistic, it had slightly less power than the scan statistic. Our results showed that the Ionita-Laza version of Kulldorff’s scan statistic had the greatest power over a range of clustering scenarios. PMID:23842950

Embedded-Based Graphics Processing Unit Cluster Platform for Multiple Sequence Alignments

PubMed Central

Wei, Jyh-Da; Cheng, Hui-Jun; Lin, Chun-Yuan; Ye, Jin; Yeh, Kuan-Yu

2017-01-01

High-end graphics processing units (GPUs), such as NVIDIA Tesla/Fermi/Kepler series cards with thousands of cores per chip, are widely applied to high-performance computing fields in a decade. These desktop GPU cards should be installed in personal computers/servers with desktop CPUs, and the cost and power consumption of constructing a GPU cluster platform are very high. In recent years, NVIDIA releases an embedded board, called Jetson Tegra K1 (TK1), which contains 4 ARM Cortex-A15 CPUs and 192 Compute Unified Device Architecture cores (belong to Kepler GPUs). Jetson Tegra K1 has several advantages, such as the low cost, low power consumption, and high applicability, and it has been applied into several specific applications. In our previous work, a bioinformatics platform with a single TK1 (STK platform) was constructed, and this previous work is also used to prove that the Web and mobile services can be implemented in the STK platform with a good cost-performance ratio by comparing a STK platform with the desktop CPU and GPU. In this work, an embedded-based GPU cluster platform will be constructed with multiple TK1s (MTK platform). Complex system installation and setup are necessary procedures at first. Then, 2 job assignment modes are designed for the MTK platform to provide services for users. Finally, ClustalW v2.0.11 and ClustalWtk will be ported to the MTK platform. The experimental results showed that the speedup ratios achieved 5.5 and 4.8 times for ClustalW v2.0.11 and ClustalWtk, respectively, by comparing 6 TK1s with a single TK1. The MTK platform is proven to be useful for multiple sequence alignments. PMID:28835734

Specialized Computer Systems for Environment Visualization

NASA Astrophysics Data System (ADS)

Al-Oraiqat, Anas M.; Bashkov, Evgeniy A.; Zori, Sergii A.

2018-06-01

The need for real time image generation of landscapes arises in various fields as part of tasks solved by virtual and augmented reality systems, as well as geographic information systems. Such systems provide opportunities for collecting, storing, analyzing and graphically visualizing geographic data. Algorithmic and hardware software tools for increasing the realism and efficiency of the environment visualization in 3D visualization systems are proposed. This paper discusses a modified path tracing algorithm with a two-level hierarchy of bounding volumes and finding intersections with Axis-Aligned Bounding Box. The proposed algorithm eliminates the branching and hence makes the algorithm more suitable to be implemented on the multi-threaded CPU and GPU. A modified ROAM algorithm is used to solve the qualitative visualization of reliefs' problems and landscapes. The algorithm is implemented on parallel systems—cluster and Compute Unified Device Architecture-networks. Results show that the implementation on MPI clusters is more efficient than Graphics Processing Unit/Graphics Processing Clusters and allows real-time synthesis. The organization and algorithms of the parallel GPU system for the 3D pseudo stereo image/video synthesis are proposed. With realizing possibility analysis on a parallel GPU-architecture of each stage, 3D pseudo stereo synthesis is performed. An experimental prototype of a specialized hardware-software system 3D pseudo stereo imaging and video was developed on the CPU/GPU. The experimental results show that the proposed adaptation of 3D pseudo stereo imaging to the architecture of GPU-systems is efficient. Also it accelerates the computational procedures of 3D pseudo-stereo synthesis for the anaglyph and anamorphic formats of the 3D stereo frame without performing optimization procedures. The acceleration is on average 11 and 54 times for test GPUs.

Embedded-Based Graphics Processing Unit Cluster Platform for Multiple Sequence Alignments.

PubMed

Wei, Jyh-Da; Cheng, Hui-Jun; Lin, Chun-Yuan; Ye, Jin; Yeh, Kuan-Yu

2017-01-01

High-end graphics processing units (GPUs), such as NVIDIA Tesla/Fermi/Kepler series cards with thousands of cores per chip, are widely applied to high-performance computing fields in a decade. These desktop GPU cards should be installed in personal computers/servers with desktop CPUs, and the cost and power consumption of constructing a GPU cluster platform are very high. In recent years, NVIDIA releases an embedded board, called Jetson Tegra K1 (TK1), which contains 4 ARM Cortex-A15 CPUs and 192 Compute Unified Device Architecture cores (belong to Kepler GPUs). Jetson Tegra K1 has several advantages, such as the low cost, low power consumption, and high applicability, and it has been applied into several specific applications. In our previous work, a bioinformatics platform with a single TK1 (STK platform) was constructed, and this previous work is also used to prove that the Web and mobile services can be implemented in the STK platform with a good cost-performance ratio by comparing a STK platform with the desktop CPU and GPU. In this work, an embedded-based GPU cluster platform will be constructed with multiple TK1s (MTK platform). Complex system installation and setup are necessary procedures at first. Then, 2 job assignment modes are designed for the MTK platform to provide services for users. Finally, ClustalW v2.0.11 and ClustalWtk will be ported to the MTK platform. The experimental results showed that the speedup ratios achieved 5.5 and 4.8 times for ClustalW v2.0.11 and ClustalWtk, respectively, by comparing 6 TK1s with a single TK1. The MTK platform is proven to be useful for multiple sequence alignments.

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.

Visual based laser speckle pattern recognition method for structural health monitoring

NASA Astrophysics Data System (ADS)

Park, Kyeongtaek; Torbol, Marco

2017-04-01

This study performed the system identification of a target structure by analyzing the laser speckle pattern taken by a camera. The laser speckle pattern is generated by the diffuse reflection of the laser beam on a rough surface of the target structure. The camera, equipped with a red filter, records the scattered speckle particles of the laser light in real time and the raw speckle image of the pixel data is fed to the graphic processing unit (GPU) in the system. The algorithm for laser speckle contrast analysis (LASCA) computes: the laser speckle contrast images and the laser speckle flow images. The k-mean clustering algorithm is used to classify the pixels in each frame and the clusters' centroids, which function as virtual sensors, track the displacement between different frames in time domain. The fast Fourier transform (FFT) and the frequency domain decomposition (FDD) compute the modal properties of the structure: natural frequencies and damping ratios. This study takes advantage of the large scale computational capability of GPU. The algorithm is written in Compute Unifies Device Architecture (CUDA C) that allows the processing of speckle images in real time.

Hardware Testing and System Evaluation: Procedures to Evaluate Commodity Hardware for Production Clusters

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

Goebel, J

2004-02-27

Without stable hardware any program will fail. The frustration and expense of supporting bad hardware can drain an organization, delay progress, and frustrate everyone involved. At Stanford Linear Accelerator Center (SLAC), we have created a testing method that helps our group, SLAC Computer Services (SCS), weed out potentially bad hardware and purchase the best hardware at the best possible cost. Commodity hardware changes often, so new evaluations happen periodically each time we purchase systems and minor re-evaluations happen for revised systems for our clusters, about twice a year. This general framework helps SCS perform correct, efficient evaluations. This article outlinesmore » SCS's computer testing methods and our system acceptance criteria. We expanded the basic ideas to other evaluations such as storage, and we think the methods outlined in this article has helped us choose hardware that is much more stable and supportable than our previous purchases. We have found that commodity hardware ranges in quality, so systematic method and tools for hardware evaluation were necessary. This article is based on one instance of a hardware purchase, but the guidelines apply to the general problem of purchasing commodity computer systems for production computational work.« less

Effective correlator for RadioAstron project

NASA Astrophysics Data System (ADS)

Sergeev, Sergey

This paper presents the implementation of programme FX-correlator for Very Long Baseline Interferometry, adapted for the project "RadioAstron". Software correlator implemented for heterogeneous computing systems using graphics accelerators. It is shown that for the task interferometry implementation of the graphics hardware has a high efficiency. The host processor of heterogeneous computing system, performs the function of forming the data flow for graphics accelerators, the number of which corresponds to the number of frequency channels. So, for the Radioastron project, such channels is seven. Each accelerator is perform correlation matrix for all bases for a single frequency channel. Initial data is converted to the floating-point format, is correction for the corresponding delay function and computes the entire correlation matrix simultaneously. Calculation of the correlation matrix is performed using the sliding Fourier transform. Thus, thanks to the compliance of a solved problem for architecture graphics accelerators, managed to get a performance for one processor platform Kepler, which corresponds to the performance of this task, the computing cluster platforms Intel on four nodes. This task successfully scaled not only on a large number of graphics accelerators, but also on a large number of nodes with multiple accelerators.

A coupled cluster theory with iterative inclusion of triple excitations and associated equation of motion formulation for excitation energy and ionization potential

NASA Astrophysics Data System (ADS)

Maitra, Rahul; Akinaga, Yoshinobu; Nakajima, Takahito

2017-08-01

A single reference coupled cluster theory that is capable of including the effect of connected triple excitations has been developed and implemented. This is achieved by regrouping the terms appearing in perturbation theory and parametrizing through two different sets of exponential operators: while one of the exponentials, involving general substitution operators, annihilates the ground state but has a non-vanishing effect when it acts on the excited determinant, the other is the regular single and double excitation operator in the sense of conventional coupled cluster theory, which acts on the Hartree-Fock ground state. The two sets of operators are solved as coupled non-linear equations in an iterative manner without significant increase in computational cost than the conventional coupled cluster theory with singles and doubles excitations. A number of physically motivated and computationally advantageous sufficiency conditions are invoked to arrive at the working equations and have been applied to determine the ground state energies of a number of small prototypical systems having weak multi-reference character. With the knowledge of the correlated ground state, we have reconstructed the triple excitation operator and have performed equation of motion with coupled cluster singles, doubles, and triples to obtain the ionization potential and excitation energies of these molecules as well. Our results suggest that this is quite a reasonable scheme to capture the effect of connected triple excitations as long as the ground state remains weakly multi-reference.

Modeling of the HiPco process for carbon nanotube production. II. Reactor-scale analysis

NASA Technical Reports Server (NTRS)

Gokcen, Tahir; Dateo, Christopher E.; Meyyappan, M.

2002-01-01

The high-pressure carbon monoxide (HiPco) process, developed at Rice University, has been reported to produce single-walled carbon nanotubes from gas-phase reactions of iron carbonyl in carbon monoxide at high pressures (10-100 atm). Computational modeling is used here to develop an understanding of the HiPco process. A detailed kinetic model of the HiPco process that includes of the precursor, decomposition metal cluster formation and growth, and carbon nanotube growth was developed in the previous article (Part I). Decomposition of precursor molecules is necessary to initiate metal cluster formation. The metal clusters serve as catalysts for carbon nanotube growth. The diameter of metal clusters and number of atoms in these clusters are some of the essential information for predicting carbon nanotube formation and growth, which is then modeled by the Boudouard reaction with metal catalysts. Based on the detailed model simulations, a reduced kinetic model was also developed in Part I for use in reactor-scale flowfield calculations. Here this reduced kinetic model is integrated with a two-dimensional axisymmetric reactor flow model to predict reactor performance. Carbon nanotube growth is examined with respect to several process variables (peripheral jet temperature, reactor pressure, and Fe(CO)5 concentration) with the use of the axisymmetric model, and the computed results are compared with existing experimental data. The model yields most of the qualitative trends observed in the experiments and helps to understanding the fundamental processes in HiPco carbon nanotube production.

Reliability Evaluation for Clustered WSNs under Malware Propagation

PubMed Central

Shen, Shigen; Huang, Longjun; Liu, Jianhua; Champion, Adam C.; Yu, Shui; Cao, Qiying

2016-01-01

We consider a clustered wireless sensor network (WSN) under epidemic-malware propagation conditions and solve the problem of how to evaluate its reliability so as to ensure efficient, continuous, and dependable transmission of sensed data from sensor nodes to the sink. Facing the contradiction between malware intention and continuous-time Markov chain (CTMC) randomness, we introduce a strategic game that can predict malware infection in order to model a successful infection as a CTMC state transition. Next, we devise a novel measure to compute the Mean Time to Failure (MTTF) of a sensor node, which represents the reliability of a sensor node continuously performing tasks such as sensing, transmitting, and fusing data. Since clustered WSNs can be regarded as parallel-serial-parallel systems, the reliability of a clustered WSN can be evaluated via classical reliability theory. Numerical results show the influence of parameters such as the true positive rate and the false positive rate on a sensor node’s MTTF. Furthermore, we validate the method of reliability evaluation for a clustered WSN according to the number of sensor nodes in a cluster, the number of clusters in a route, and the number of routes in the WSN. PMID:27294934

Reliability Evaluation for Clustered WSNs under Malware Propagation.

PubMed

Shen, Shigen; Huang, Longjun; Liu, Jianhua; Champion, Adam C; Yu, Shui; Cao, Qiying

2016-06-10

We consider a clustered wireless sensor network (WSN) under epidemic-malware propagation conditions and solve the problem of how to evaluate its reliability so as to ensure efficient, continuous, and dependable transmission of sensed data from sensor nodes to the sink. Facing the contradiction between malware intention and continuous-time Markov chain (CTMC) randomness, we introduce a strategic game that can predict malware infection in order to model a successful infection as a CTMC state transition. Next, we devise a novel measure to compute the Mean Time to Failure (MTTF) of a sensor node, which represents the reliability of a sensor node continuously performing tasks such as sensing, transmitting, and fusing data. Since clustered WSNs can be regarded as parallel-serial-parallel systems, the reliability of a clustered WSN can be evaluated via classical reliability theory. Numerical results show the influence of parameters such as the true positive rate and the false positive rate on a sensor node's MTTF. Furthermore, we validate the method of reliability evaluation for a clustered WSN according to the number of sensor nodes in a cluster, the number of clusters in a route, and the number of routes in the WSN.

Hierarchical Aligned Cluster Analysis for Temporal Clustering of Human Motion.

PubMed

Zhou, Feng; De la Torre, Fernando; Hodgins, Jessica K

2013-03-01

Temporal segmentation of human motion into plausible motion primitives is central to understanding and building computational models of human motion. Several issues contribute to the challenge of discovering motion primitives: the exponential nature of all possible movement combinations, the variability in the temporal scale of human actions, and the complexity of representing articulated motion. We pose the problem of learning motion primitives as one of temporal clustering, and derive an unsupervised hierarchical bottom-up framework called hierarchical aligned cluster analysis (HACA). HACA finds a partition of a given multidimensional time series into m disjoint segments such that each segment belongs to one of k clusters. HACA combines kernel k-means with the generalized dynamic time alignment kernel to cluster time series data. Moreover, it provides a natural framework to find a low-dimensional embedding for time series. HACA is efficiently optimized with a coordinate descent strategy and dynamic programming. Experimental results on motion capture and video data demonstrate the effectiveness of HACA for segmenting complex motions and as a visualization tool. We also compare the performance of HACA to state-of-the-art algorithms for temporal clustering on data of a honey bee dance. The HACA code is available online.

Applying a 2D based CAD scheme for detecting micro-calcification clusters using digital breast tomosynthesis images: an assessment

NASA Astrophysics Data System (ADS)

Park, Sang Cheol; Zheng, Bin; Wang, Xiao-Hui; Gur, David

2008-03-01

Digital breast tomosynthesis (DBT) has emerged as a promising imaging modality for screening mammography. However, visually detecting micro-calcification clusters depicted on DBT images is a difficult task. Computer-aided detection (CAD) schemes for detecting micro-calcification clusters depicted on mammograms can achieve high performance and the use of CAD results can assist radiologists in detecting subtle micro-calcification clusters. In this study, we compared the performance of an available 2D based CAD scheme with one that includes a new grouping and scoring method when applied to both projection and reconstructed DBT images. We selected a dataset involving 96 DBT examinations acquired on 45 women. Each DBT image set included 11 low dose projection images and a varying number of reconstructed image slices ranging from 18 to 87. In this dataset 20 true-positive micro-calcification clusters were visually detected on the projection images and 40 were visually detected on the reconstructed images, respectively. We first applied the CAD scheme that was previously developed in our laboratory to the DBT dataset. We then tested a new grouping method that defines an independent cluster by grouping the same cluster detected on different projection or reconstructed images. We then compared four scoring methods to assess the CAD performance. The maximum sensitivity level observed for the different grouping and scoring methods were 70% and 88% for the projection and reconstructed images with a maximum false-positive rate of 4.0 and 15.9 per examination, respectively. This preliminary study demonstrates that (1) among the maximum, the minimum or the average CAD generated scores, using the maximum score of the grouped cluster regions achieved the highest performance level, (2) the histogram based scoring method is reasonably effective in reducing false-positive detections on the projection images but the overall CAD sensitivity is lower due to lower signal-to-noise ratio, and (3) CAD achieved higher sensitivity and higher false-positive rate (per examination) on the reconstructed images. We concluded that without changing the detection threshold or performing pre-filtering to possibly increase detection sensitivity, current CAD schemes developed and optimized for 2D mammograms perform relatively poorly and need to be re-optimized using DBT datasets and new grouping and scoring methods need to be incorporated into the schemes if these are to be used on the DBT examinations.

Method and system for data clustering for very large databases

NASA Technical Reports Server (NTRS)

Livny, Miron (Inventor); Zhang, Tian (Inventor); Ramakrishnan, Raghu (Inventor)

1998-01-01

Multi-dimensional data contained in very large databases is efficiently and accurately clustered to determine patterns therein and extract useful information from such patterns. Conventional computer processors may be used which have limited memory capacity and conventional operating speed, allowing massive data sets to be processed in a reasonable time and with reasonable computer resources. The clustering process is organized using a clustering feature tree structure wherein each clustering feature comprises the number of data points in the cluster, the linear sum of the data points in the cluster, and the square sum of the data points in the cluster. A dense region of data points is treated collectively as a single cluster, and points in sparsely occupied regions can be treated as outliers and removed from the clustering feature tree. The clustering can be carried out continuously with new data points being received and processed, and with the clustering feature tree being restructured as necessary to accommodate the information from the newly received data points.

Cognitive Model Exploration and Optimization: A New Challenge for Computational Science

DTIC Science & Technology

2010-01-01

Introduction Research in cognitive science often involves the generation and analysis of computational cognitive models to explain various...HPC) clusters and volunteer computing for large-scale computational resources. The majority of applications on the Department of Defense HPC... clusters focus on solving partial differential equations (Post, 2009). These tend to be lean, fast models with little noise. While we lack specific

A Fast Implementation of the ISOCLUS Algorithm

NASA Technical Reports Server (NTRS)

Memarsadeghi, Nargess; Mount, David M.; Netanyahu, Nathan S.; LeMoigne, Jacqueline

2003-01-01

Unsupervised clustering is a fundamental tool in numerous image processing and remote sensing applications. For example, unsupervised clustering is often used to obtain vegetation maps of an area of interest. This approach is useful when reliable training data are either scarce or expensive, and when relatively little a priori information about the data is available. Unsupervised clustering methods play a significant role in the pursuit of unsupervised classification. One of the most popular and widely used clustering schemes for remote sensing applications is the ISOCLUS algorithm, which is based on the ISODATA method. The algorithm is given a set of n data points (or samples) in d-dimensional space, an integer k indicating the initial number of clusters, and a number of additional parameters. The general goal is to compute a set of cluster centers in d-space. Although there is no specific optimization criterion, the algorithm is similar in spirit to the well known k-means clustering method in which the objective is to minimize the average squared distance of each point to its nearest center, called the average distortion. One significant feature of ISOCLUS over k-means is that clusters may be merged or split, and so the final number of clusters may be different from the number k supplied as part of the input. This algorithm will be described in later in this paper. The ISOCLUS algorithm can run very slowly, particularly on large data sets. Given its wide use in remote sensing, its efficient computation is an important goal. We have developed a fast implementation of the ISOCLUS algorithm. Our improvement is based on a recent acceleration to the k-means algorithm, the filtering algorithm, by Kanungo et al.. They showed that, by storing the data in a kd-tree, it was possible to significantly reduce the running time of k-means. We have adapted this method for the ISOCLUS algorithm. For technical reasons, which are explained later, it is necessary to make a minor modification to the ISOCLUS specification. We provide empirical evidence, on both synthetic and Landsat image data sets, that our algorithm's performance is essentially the same as that of ISOCLUS, but with significantly lower running times. We show that our algorithm runs from 3 to 30 times faster than a straightforward implementation of ISOCLUS. Our adaptation of the filtering algorithm involves the efficient computation of a number of cluster statistics that are needed for ISOCLUS, but not for k-means.

Hartree-Fock calculation of the differential photoionization cross sections of small Li clusters

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

Galitskiy, S. A.; Artemyev, A. N.; Jänkälä, K.

2015-01-21

Cross sections and angular distribution parameters for the single-photon ionization of all electron orbitals of Li{sub 2−8} are systematically computed in a broad interval of the photoelectron kinetic energies for the energetically most stable geometry of each cluster. Calculations of the partial photoelectron continuum waves in clusters are carried out by the single center method within the Hartree-Fock approximation. We study photoionization cross sections per one electron and analyze in some details general trends in the photoionization of inner and outer shells with respect to the size and geometry of a cluster. The present differential cross sections computed for Li{submore » 2} are in a good agreement with the available theoretical data, whereas those computed for Li{sub 3−8} clusters can be considered as theoretical predictions.« less

Operating Dedicated Data Centers - Is It Cost-Effective?

NASA Astrophysics Data System (ADS)

Ernst, M.; Hogue, R.; Hollowell, C.; Strecker-Kellog, W.; Wong, A.; Zaytsev, A.

2014-06-01

The advent of cloud computing centres such as Amazon's EC2 and Google's Computing Engine has elicited comparisons with dedicated computing clusters. Discussions on appropriate usage of cloud resources (both academic and commercial) and costs have ensued. This presentation discusses a detailed analysis of the costs of operating and maintaining the RACF (RHIC and ATLAS Computing Facility) compute cluster at Brookhaven National Lab and compares them with the cost of cloud computing resources under various usage scenarios. An extrapolation of likely future cost effectiveness of dedicated computing resources is also presented.

Coupled Cluster Method with Single and Double Excitations Tailored by Matrix Product State Wave Functions.

PubMed

Veis, Libor; Antalík, Andrej; Brabec, Jiří; Neese, Frank; Legeza, Örs; Pittner, Jiří

2016-10-03

In the past decade, the quantum chemical version of the density matrix renormalization group (DMRG) method has established itself as the method of choice for calculations of strongly correlated molecular systems. Despite its favorable scaling, it is in practice not suitable for computations of dynamic correlation. We present a novel method for accurate "post-DMRG" treatment of dynamic correlation based on the tailored coupled cluster (CC) theory in which the DMRG method is responsible for the proper description of nondynamic correlation, whereas dynamic correlation is incorporated through the framework of the CC theory. We illustrate the potential of this method on prominent multireference systems, in particular, N 2 and Cr 2 molecules and also oxo-Mn(Salen), for which we have performed the first post-DMRG computations in order to shed light on the energy ordering of the lowest spin states.

Numerical study of base pressure characteristic curve for a four-engine clustered nozzle configuration

NASA Technical Reports Server (NTRS)

Wang, Ten-See

1993-01-01

The objective of this study is to benchmark a four-engine clustered nozzle base flowfield with a computational fluid dynamics (CFD) model. The CFD model is a three-dimensional pressure-based, viscous flow formulation. An adaptive upwind scheme is employed for the spatial discretization. The upwind scheme is based on second and fourth order central differencing with adaptive artificial dissipation. Qualitative base flow features such as the reverse jet, wall jet, recompression shock, and plume-plume impingement have been captured. The computed quantitative flow properties such as the radial base pressure distribution, model centerline Mach number and static pressure variation, and base pressure characteristic curve agreed reasonably well with those of the measurement. Parametric study on the effect of grid resolution, turbulence model, inlet boundary condition and difference scheme on convective terms has been performed. The results showed that grid resolution had a strong influence on the accuracy of the base flowfield prediction.

Configuration and Management of a Cluster Computing Facility in Undergraduate Student Computer Laboratories

ERIC Educational Resources Information Center

Cornforth, David; Atkinson, John; Spennemann, Dirk H. R.

2006-01-01

Purpose: Many researchers require access to computer facilities beyond those offered by desktop workstations. Traditionally, these are offered either through partnerships, to share the cost of supercomputing facilities, or through purpose-built cluster facilities. However, funds are not always available to satisfy either of these options, and…