基于CMP多核集群的混合并行编程技术研究

计算机科学 ›› 2014, Vol. 41 ›› Issue (2): 19-22.

基于CMP多核集群的混合并行编程技术研究

王文义,王春霞,王杰

中原工学院并行处理技术研究所郑州450007;中原工学院并行处理技术研究所郑州450007;中原工学院并行处理技术研究所郑州450007

出版日期:2018-11-14 发布日期:2018-11-14
基金资助:
本文受国家863计划项目(2008AA01A315),河南省基础与前沿技术研究项目(122300410314)资助

Research on Hybrid Parallel Programming Technique Based on CMP Multi-cure Cluster

WANG Wen-yi,WANG Chun-xia and WANG Jie

Online:2018-11-14 Published:2018-11-14

摘要/Abstract

摘要： 高性能科学计算 (High Performance Science Computing,简称HPC)是验证某些理论和测试计算机系统处理能力的一种有效的实验手段。鉴于目前CMP(Chip Multi-processor)多核集群已变得越来越普及,尝试对由MPI和OpenMP两种不同并行编程技术构成的混合编程模式做一些实验性的研究。通过对程序执行时间和加速比的实验数据分析,可以看出在多核和多节点集群上采用细粒度的混合并行编程方法较单一使用MPI并行编程方法更加合理和高效,也更能体现出系统硬软件的特性与优势。

关键词: 高性能计算,CMP多核集群,墙钟时间,MPI+OpenMP混合并行编程中图法分类号TP39文献标识码A

Abstract: When validating some theory and testing computer systems processing capacity,high-performance scientific computing is an effective experimental means．Currently,as CMP multi-core clusters become increasingly common,this article attempted to do some experimental studies to the MPI and OpenMP two different parallel programming technique consisting of hybrid programming model．Through analyzing the experimental data of the program execution time and the speedup,it can be seen that in multi-core and multi-node clusters using fine-grained Hybrid parallel programming method than the single using MPI parallel programming method will be more rational and efficient,so it can also better reflect the features and advantages of system hardware and software.

Key words: High performance computing,CMP multi-core cluster,Wall clock time,Hybrid parallel programming with MPICH and OpenMP

王文义,王春霞,王杰. 基于CMP多核集群的混合并行编程技术研究[J]. 计算机科学, 2014, 41(2): 19-22. https://doi.org/

WANG Wen-yi,WANG Chun-xia and WANG Jie. Research on Hybrid Parallel Programming Technique Based on CMP Multi-cure Cluster[J]. Computer Science, 2014, 41(2): 19-22. https://doi.org/

参考文献

[1] Hwang K．Advanced Computer Architecture:Parallelism Sca-lability Programmability[M]．New York:McGraw-Hill Inc.,1993
[2] Group W,Skjellum E L A．Using MPI——Portable ParallelProgramming with the Massage Interface(Second Edition)[M].CambridgeMassachusetts,London,England:The MITPress,1999
[3] 都志辉．高性能计算之并行编程技术MPI并行程序设计[M].北京:清华大学出版社,2001
[4] Robert S A J.Multi-core Programming:Increasing performance Through Software Multi-threading [M].李宝峰,富弘毅,李韬,译.北京:电子工业出版社,2007:145-283
[5] Chandra R,Dagum L,Kohr D,et al.Menon:Parallel programming in OpenMP[M]．Morgan Kaufmann Publisher,Inc.,San Francisco,CA,USA,2001
[6] OpenMP C and C⁺⁺ Application Program Interface．version3.May 2008.http://www．openmp.org
[7] Brown R.Performance and Productivity Comparison BetweenOpenMP and MPI[J].Int Parallel Prog,2007,5:441-458
[8] 章隆兵,吴少刚,蔡飞.适合集群OpenMP系统的制导扩展[J].计算机学报,2004,7(8):1129-1135
[9] 陈永健.OpenMP编译与优化技术研究[D].北京:清华大学,2004
[10] Core i7QPI技术解密．http://wenku.baidu.com/view/63e77d160b4e767f5acfcea -e.html.Wang D T.The CELL microprocessor.Real World Technologies,2005
[11] Smith L,Bull M.Development of mixed mode MPI+OpenMP applications [J]．Scientific Programming,2001,9:83-98
[12] Lusk E,Chan A.Early Experiments with the OpenMP/MPIHybrid Programming Model [C]∥IWOMP’08Proceedings of the 4th International Conference on Open MP in a New era of Parallelism.Springer,2008,4:36-47
[13] 叶晓敏．基于多核处理器并行EDA算法研究[D].复旦大学

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed