微信服务号

微信订阅号

2025-4-25- 10
Home > Archive>Volume 15, Issue 6, 2004 >842-849
PDF HTML XML Export Cite reminder
Shared-Memory Versus Message-Passing on PC Cluster
Author:
Affiliation:

  • Article
    • | |
  • Metrics
    • |
  • Reference [16]
    • |
  • Related [20]
    • |
  • Cited by
    • | |
  • Comments
    Abstract:

    Two parallel programming models of shared-memory and message-passing are widely adopted. The programmability of message-passing is poor, while that of shared-memory is good. The OpenMP Application Programming Interface is an emerging standard for shared-memory. OpenMP on cluster supplies an OpenMP computing environment on cluster of workstations or PCs, which combines the friendly programmability of shared-memory with the fine scalability of cluster. Taking 7 well-known parallel applications on a cluster of PCs, this paper compares the performance of OpenMP/JIAJIA, an OpenMP system on cluster, with that of MPI, a typical message passing system. Experimental results show that the performance of OpenMP is averagely equal to 81% of MPI for the 7 applications running on 8-nodes, but the former is easier to use than the latter.

    Reference
    [1]Snir M, Otto S, Huss-Lederman S, Walker D, Dongarra J. MPI: The Complete Reference. London: MIT Press, 1996.
    [2]OpenMP Architecture Review Board. OpenMP C and C++ Application Program Interface, Version 2.0. 2002.
    [3]Lu HH, Hu YC, Zwaenepoel W. OpenMP on networks of workstations. In: Benton V, ed. Proc. of the Supercomputing'98. Orlando:IEEE Computer Society. 1998. 1～15
    [4]Sato M, Sato S, Kusano K, Tanaka Y. Design of OpenMP compiler for an SMP cluster. In: Proc. of the 1st European Workshop on OpenMP. 1999. 32～39.
    [5]Brunschen C, Brorsson M. OdinMP/CCP-A portable implementation of OpenMP for C. Concurrency and Computation: Practice and Experience, 2000,12(12):1193～1203.
    [6]Lu HH, Dwarkadas S, Cox AL, Zwaenepoel W. Message passing versus distributed shared memory on networks of workstations. In:Redelfs A, ed. Proc. of the 1995 ACM/IEEE Supercomputing Conf. San Diego: ACM/IEEE Computer Society, 1995. 865～906
    [7]Lu HH, Dwarkadas S, Cox AL, Zwaenepoel W. Quantifying the performance difference between PVM and Treadmarks. Journal of Parallel and Distributed Computation, 1997,43(2):65～78.
    [8]Tang ZM, Shi WS, Hu WW. Message-Passing versus shared-memory on dawning 1000A. Chinese Journal of Computers, 2000,23(2):134～140 (in Chinese with English abstract).
    [9]Hu MC, Shi G, Hu WW, Tang ZM, Zhang FX. Comparing JIAJIA with MPI on PC cluster. Journal of Software, 2003,14(7):1187～1194 ( in Chinese with English abstract). http://www.jos.org.cn/1000-9825/14/1187.htm
    [10]Hu WW, Shi WS, Tang ZM. JIAJIA: A software DSM system based on a new cache coherence protocol. In: Sloot PMA, Bubak M,Hoekstra AG, Hertzberger B, eds. Proc. of the HPCN Europe'99. LNCS 1593, Amsterdam, 1999. 463～472.
    [11]Amarasinghe SP, Anderson JM, Lam MS, Tseng CW. An overview of the SUIF compiler for scalable parallel machines. In: David H, ed. Proc. of the 7th SIAM Conf. on Parallel Processing for Scientific Computing. Philadelphia: SIAM, 1995. 662～667.
    [12]Bailey D, Harris T, Saphir W, van der Wijngaart R, Woo A, Yarrow M. The NAS parallel benchmarks 2.0. Technical Report,NAS-95-020, 1995.
    [13]Woo SC, Ohara M, Torrie E, Singh JP, Gupta A. The SPLASH-2 programs: Characterization and methodological considerations. In:Proc. of the 22nd Annual Symp. on Computer Architecture. New York: ACM Press, 1995.24～36.
    [14]Aslot V, Domeika M, Eigenmann R, et al. SPEComp: A new benchmark suite for measuring parallel computer performance. In:E igenmann, Michael JV, eds. Proc. of the Workshop on OpenMP Application and Tools. LNCS 2104, 2001. 1～ 10.
    [15]唐志敏,施巍松,胡伟武.曙光1000A上消息传递与共享存储的比较.计算机学报,2000,23(2):134～140.
    [16]胡明昌,史岗,胡伟武,唐志敏,张福新.PC机群上JIAJIA与MPI的比较.软件学报,2003,14(7):1187～1194.http://www.jos.org.cn/1000-9825/14/1187.htm
    Cited by
    Comments
    Comments
    分享到微博
    Submit
Get Citation

章隆兵,吴少刚,蔡飞,胡伟武. PC机群上共享存储与消息传递的比较.软件学报,2004,15(6):842-849

Copy
Share
Article Metrics
  • Abstract:4455
  • PDF: 5657
  • HTML: 0
  • Cited by: 0
History
  • Received:June 01,2004
Links:Institute of Software, CASChina Computer FederationCASNSFC
You are the first2038604Visitors
Copyright: Institute of Software, Chinese Academy of Sciences Beijing ICP No. 05046678-4
Address：4# South Fourth Street, Zhong Guan Cun, Beijing 100190,Postal Code：100190
Phone：010-62562563 Fax：010-62562533 Email：jos@iscas.ac.cn
Technical Support：Beijing Qinyun Technology Development Co., Ltd.

Beijing Public Network Security No. 11040202500063