2009 DoD High Performance Computing Modernization Program Users Group Conference最新文献_第6页

Measuring and Understanding Variation in Benchmark Performance 测量和理解基准性能的变化

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.72

N. Wright, Shava Smallen, C. Olschanowsky, J. Hayes, A. Snavely

{"title":"Measuring and Understanding Variation in Benchmark Performance","authors":"N. Wright, Shava Smallen, C. Olschanowsky, J. Hayes, A. Snavely","doi":"10.1109/HPCMP-UGC.2009.72","DOIUrl":"https://doi.org/10.1109/HPCMP-UGC.2009.72","url":null,"abstract":"Runtime irreproducibility complicates application performance evaluation on today’s high performance computers. Performance can vary significantly between seemingly identical runs; this presents a challenge to benchmarking as well as a user, who is trying to determine whether the change they made to their code is an actual improvement. In order to gain a better understanding of this phenomenon, we measure the runtime variation of two applications, PARAllel Total Energy Code (PARATEC) and Weather Research and Forecasting (WRF), on three different machines. Key associated metrics are also recorded. The data is then used to 1) quantify the magnitude and distribution of the variations and 2) gain an understanding as why the variations occur. Using our lightweight framework, Integrated Performance Monitoring (IPM), to understand the performance characteristics of individual runs, and the Inca framework to automate the procedure measurements were collected over a month’s time. The results indicate that performance can vary up to 25% and is almost always due to contention for network resources. We also found that the variation differs between machines and is almost always greater on machines with lower performing networks.","PeriodicalId":268639,"journal":{"name":"2009 DoD High Performance Computing Modernization Program Users Group Conference","volume":"408 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116242530","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 27

Full Annulus High Fidelity Fan and Compressor Simulations 全环空高保真风扇和压缩机仿真

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.13

S. Gorrell, Jixian Yao, Michael G. List

引用次数: 2

Using Star-P® on DoD High Performance Computing Systems 在国防部高性能计算系统上使用Star-P®

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.64

B. Elton, S. Samsi, H. Smith, L. Humphrey, S. Ahalt, A. Chalker, Niraj Srivastava, A. H. Abdullah, P. Boyle

引用次数: 1

Quantum Lattice-Gas Algorithm for Quantum Turbulence - CAP Simulations on 12,288 Cores of Cray XT-5 Einstein at NAVO 量子湍流的量子点阵-气体算法-在NAVO的12288个Cray XT-5爱因斯坦核上的CAP模拟

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.20

G. Vahala, J. Yepez, M. Soe, L. Vahala, S. Ziegeler

引用次数: 2

Electronic Properties of High-Performance Capacitor Materials and Nanoscale Multiterminal Devices 高性能电容器材料和纳米多终端器件的电子特性

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.51

J. Bernholc, L. Yu, V. Ranjan, M. Nardelli, W. Lu, K. Saha, V. Meunier

{"title":"Electronic Properties of High-Performance Capacitor Materials and Nanoscale Multiterminal Devices","authors":"J. Bernholc, L. Yu, V. Ranjan, M. Nardelli, W. Lu, K. Saha, V. Meunier","doi":"10.1109/HPCMP-UGC.2009.51","DOIUrl":"https://doi.org/10.1109/HPCMP-UGC.2009.51","url":null,"abstract":"Recent advances in theoretical methods combined with the advent of massively-parallel supercomputers allow one to reliably simulate the properties of complex materials and device structures from first principles. We describe applications in two general areas: (i) novel ferroelectric oxide-polymer composites for ultrahigh power density capacitors, necessary for pulsed power applications, such as electric discharges, power conditioning, and dense electronic circuitry, and (ii) electron transport properties of ballistic, multi-terminal molecular devices, which could form the basis for ultraspeed electronics and spintronics. For capacitor materials, we investigate the dielectric properties of PbTiO3 slabs and polypropylene/PbTiO3 nanocomposites. We evaluate both the optical and static local dielectric permittivity profiles for isolated PbTiO3 slabs and across the polypropylene/PbTiO3 interface. For thin ferroelectric slabs, we find that in order to maintain the ferroelectric structure, it is necessary to introduce compensating surface charges. Our results show that: (i) the surface-and interface-induced modifications to dielectric permittivity in polymer/metal-oxide composites are localized to only a few atomic layers; (ii) the interface effects are mainly confined to the metal-oxide side; and (iii) metal-oxide particles larger than a few nanometers retain the average macroscopic value of bulk dielectric permittivity. Turning to nanoelectronic devices, we investigate ballistic electron transport through a paradigmatic four-terminal molecular electronic device. In contrast to a conventional two-terminal setup, the same organic molecule placed between four electrodes exhibits new properties, such as a pronounced negative differential resistance.","PeriodicalId":268639,"journal":{"name":"2009 DoD High Performance Computing Modernization Program Users Group Conference","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132777581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Realization of Linear Wave-Propagation Models from HPC Simulations 基于HPC仿真的线性波传播模型的实现

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.57

S. Ketcham, M. Parker, M. Phan

{"title":"Realization of Linear Wave-Propagation Models from HPC Simulations","authors":"S. Ketcham, M. Parker, M. Phan","doi":"10.1109/HPCMP-UGC.2009.57","DOIUrl":"https://doi.org/10.1109/HPCMP-UGC.2009.57","url":null,"abstract":"Modeling of sound propagation in complex environments requires high performance computing (HPC) to simulate three-dimensional wave fields with realistic fidelity. This is especially true for urban areas, where sound waves reflect and diffract due to the built-up infrastructure. HPC can predict these wave fields with desired fidelity, but the computational investment would have greater return if reduced-size models that operate with considerably less computational resources could be produced from the results. The objective of this work is to develop such models. The work applies a modified version of the Eigensystem Realization Algorithm, using Markov parameters from HPC input-output response functions, to generate state-space models that simulate hundreds of thousands of output signals of the HPC wave field. The results include predicted acoustic signals and signatures from realized models, using a source with a different time series than the source used to generate the Markov parameters. We compare wave-field signals from reduced-order models with HPC model signals over a large urban domain, adjusting the model order and accuracy by singular-value cutoff. We conclude that the method produces efficient high-fidelity models of sound propagation in complex environments.","PeriodicalId":268639,"journal":{"name":"2009 DoD High Performance Computing Modernization Program Users Group Conference","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117228536","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Multi-scale Forecasting and Targeting of Tropical Cyclones in the Western Pacific 西太平洋热带气旋的多尺度预报与瞄准

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.44

J. Doyle, C. Reynolds, Hao Jin, R. Hodur

引用次数: 0

High Performance Computational Modeling of Unsteady Surface Loads in Complex Weapons Bays 复杂武器舱非定常表面载荷的高性能计算模型

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.14

S. Arunajatesan, N. Sinha, M. Stanek, J. Grove, Rudy A. Johnson

引用次数: 4

Comparison of Empirical and Theoretical Computations of Velocity for a Cold Spray Nozzle 冷喷嘴速度经验计算与理论计算的比较

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.10

S. Dinavahi, V. Champagne, D. Helfritch

{"title":"Comparison of Empirical and Theoretical Computations of Velocity for a Cold Spray Nozzle","authors":"S. Dinavahi, V. Champagne, D. Helfritch","doi":"10.1109/HPCMP-UGC.2009.10","DOIUrl":"https://doi.org/10.1109/HPCMP-UGC.2009.10","url":null,"abstract":"Cold spray is a process whereby micron-size particles are accelerated to high velocity through entrainment in a gas undergoing expansion in a rocket nozzle and are subsequently impacted upon a surface. The impacted particles, which can be combinations of metals, ceramics and polymeric materials, form a consolidated structure that can be several centimeters thick. The characteristics of this structure depend on the initial characteristics of the metal powder and upon the impact velocity. Two-dimensional axi-symmetric computations of the flow through a converging, diverging nozzle were performed using the Reynolds-Averaged Navier-Stokes (RANS) code, Computational Fluid Dynamics++ (CFD++), on the Army Research Laboratory, Department of Defense (DoD) Supercomputing Resource Center (ARL DSRC) computers. Aluminum particles of constant diameter were injected at the entrance of a De Laval converging, diverging nozzle. The Eulerian Disperse Phase (EDP) capability in CFD++ was used for these simulations. The EDP model couples the dispersed phase with the fluid dynamics. In addition, onedimensional (1D), isentropic, gas-dynamic equations were solved for the same geometry and initial conditions. The results from the RANS computations and 1D calculation compared favorably, considering the difference in governing equations.","PeriodicalId":268639,"journal":{"name":"2009 DoD High Performance Computing Modernization Program Users Group Conference","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131817409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

An Analysis of GPU Parallel Computing GPU并行计算分析

2009 DoD High Performance Computing Modernization Program Users Group Conference Pub Date : 2009-06-15 DOI: 10.1109/HPCMP-UGC.2009.59

S. Park

引用次数: 6