作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2008, Vol. 34 ›› Issue (6): 80-81. doi: 10.3969/j.issn.1000-3428.2008.06.029

• 软件技术与数据库 • 上一篇    下一篇

有限元单元计算子程序的OpenMP并行化

宋 刚1,2,3,蒋孟奇1,2,3,张云泉2,3,李玉成2   

  1. (1. 中国科学院研究生院,北京 100080;2. 中国科学院软件研究所并行计算实验室,北京 100080;3. 中国科学院计算机科学国家重点实验室,北京 100080)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2008-03-20 发布日期:2008-03-20

OpenMP Parallelization of Element Computation Subroutine of Finite Element Method

SONG Gang1,2,3, JIANG Meng-qi1,2,3, ZHANG Yun-quan2,3, LI Yu-cheng2   



  1. (1. Graduate University of Chinese Academy of Sciences, Beijing 100080; 2. Parallel Computing Lab, Institute of Software, Chinese Academy of Sciences, Beijing 100080; 3. State Key Lab of Computer Science, Chinese Academy of Sciences, Beijing 100080)
  • Received:1900-01-01 Revised:1900-01-01 Online:2008-03-20 Published:2008-03-20

摘要:

Intel和AMD双核乃至4核处理器的推出,使得并行计算已经普及到PC机。为了充分利用多核,需要对原有程序进行多线程改造,使其充分利用多核处理带来的性能提升。该文利用共享存储编程的工业标准OpenMP对有限元方法涉及的单元计算子程序进行了并行化实现。在机群的一个双CPU的SMP节点上的测试表明,共享并行化使得该单元子程序的性能提高了一倍。

关键词: 并行编程, 多线程, 多核, 有限元

Abstract: With the release of the Intel and AMD’s dual-core and muti-core processor, parallel computing becomes pervasive. To make effective use of the muti-core processor, the original sequential program has to be parallelized. In this paper, the OpenMP is used to parallelize the element computation subroutine of finite element method. And the test on a SMP node with dual-CPU of a cluster shows that the performance is improved almost twice compared with the original sequential program.

Key words: parallel programming, multithread, multi-core, finite element

中图分类号: