登录 | 注册 | 充值 | 退出 | 公司首页 | 繁体中文 | 满意度调查
综合馆
一个面向I/O密集型并行应用的性能模型
  • 摘要

    近几年,性能模型作为一种新的并行系统性能分析方法,得到学术界和工业界的广泛重视.给出了一个开放式性能模型框架结构PMPS(n)并实现了该框架下的一个面向I/O密集型并行应用模型PMPS(3),使用该模型分析了各种NPB程序在PⅣ机群系统上的性能.实验结果表明,对存储密集型应用,PMPS(3)模型与PERC模型预测结果相当;对I/O密集型应用性能的预测,PMPS(3)模型优于PERC模型.进一步分析发现,应用的数据相关、控制相关和操作重叠会影响模型预测结果.实验结果还说明了PMPS(n)性能模型具有很好的扩展性.

  • 作者

    陈永然  齐星云  窦文华  Chen Yongran  Qi Xingyun  Dou Wenhua 

  • 作者单位

    国防科学技术大学计算机学院,长沙,410073

  • 刊期

    2007年4期 ISTIC EI PKU

  • 关键词

    性能模型  机器轮廓  应用特征  褶合方法  I/O密集型  并行系统 

参考文献
  • [1] 袁伟,张云泉,孙家昶,李玉成. 国产万亿次机群系统NPB性能测试分析. 计算机研究与发展, 2005,6
  • [2] 迟利华,刘杰,胡庆丰. 数值并行计算可扩展性评价与测试. 计算机研究与发展, 2005,6
  • [3] NERSC;The Performance Evaluation Research Center. . http://perc.nersc.gov, 2005-12-16
  • [4] Intel Pin2.0. http://rogue.colorado.edu/pin/, 2005-03-02
  • [5] SDSC. MAPS. http://www.sdsc.edu/pmac/MAPs/maps.html, 2005-08-21
  • [6] SDSC. MetaSim. http://www.sdsc.edu/pmac/MetaSim/metasim.html, 2005-08-21
  • [7] Pallas. Pallas MPI benchmarks. http://www.pallas.com/e/products/index.htm, 2005-12-24
  • [8] A Snavely;N Wolter;L Carrington. Modeling application performance by convolving machine signatures with application profiles. Austin, 2001
  • [9] L Carrington;Gao Xiaofeng;N Wolter. Performance sensitivity studies for strategic applications. Nashville, 2005
  • [10] D Kotz;N Nieuwejaar. Flexibility and performance of parallel file system. ACM Operating Systems Review, 1996,02
  • [11] University of Miami. Hybrid coordinate ocean model. http://oceanmodeling.rsmas.miami.edu/hycom/, 2005-04-03
  • [12] SISC. I/O bench. http://www.sdsc.edu/pmac/Benchmark/iobench/, 2005-08-21
  • [13] H David;S Bronis. Performance research:Current status and future directions. http://perc.nersc.gov/main.htm, 2005
  • [14] A Snavely;L Carrington;N Wolter. A framework for performance modeling and prediction. Baltimore, 2002
  • [15] L Carrington;N Wolter;A Snavely. Applying an automated framework to produce accurate blind performance predictions of full-scale HPC applications. Williamsburg, 2004
  • [16] NERSC. CEPBA. http://www.nersc.gov/nusers/resources/software/tools/cepba.php, 2005-12-24
  • [17] R Ross;D Nurmi;A Cheng. A case study in application I/O on Linux clusters. Denver, 2001
  • [18] PARL. The parallel virtual file system. http://www.parl.clemson.edu/pvfs/index.html, 2004-10-04
  • [19] P H Carns;W B Ligon Ⅲ;R B Ross. PVFS:A parallel file system for Linux clusters. Atlanta, 2000
  • [20] Rajeev Thakur. ROMIO:A high-performance portable MPI-IO implementation. http://www-unix.mcs.anl.gov/romio/, 2003-07-23
  • [21] Rajeev Thakur;William Gropp;Ewing Lusk. On implementing MPI-IO portably and with high performance. Atlanta, 1999
  • [22] Los Alamos National Security. Los Alamos National Laboratory (LANL). http://www.lanl.gov, 2005-12-17
  • [23] The HDF Group. Hierarchical data format 5. http://hdf.ncsa.uiuc.edu/, 2005-12-24
  • [24] Ryan Spaulding. NAS parallel benchmarks. http://www.nas.nasa.gov/software/npb, 2006-01-01
  • [25] F Rob;Van der Wijngaart. NAS parallel benchmark I/O version 2.4[Tech Rep:NAS-03-002]. NAS, 2003
查看更多︾
相似文献 查看更多>>
54.90.204.233