1st International ICST Conference on Scalable Information Systems

Research Article

Automatic code generation of data decomposition

  • @INPROCEEDINGS{10.1145/1146847.1146859,
        author={Ya-Nan  Shen and Rong-Cai  Zhao and Jian-Min  Pang},
        title={Automatic code generation of data decomposition},
        proceedings={1st International ICST Conference on Scalable Information Systems},
        keywords={parallelizing compilers messagepassing linear inequalities data decomposition},
  • Ya-Nan Shen
    Rong-Cai Zhao
    Jian-Min Pang
    Year: 2006
    Automatic code generation of data decomposition
    DOI: 10.1145/1146847.1146859
Ya-Nan Shen1,2,*, Rong-Cai Zhao1,*, Jian-Min Pang1,*
  • 1: ZhengZhou Information Science and Technology Institute ZhengZhou 450002, China
  • 2: .
*Contact email: sssyyynnn@hotmail.com, zrc001@371.net, jianmin_pang@hotmail.com


How to decompose or map data of programs automatically onto scalable parallel processors is a key issue in developing parallelizing compilers in DSM architecture. Data locality is crucial for parallelized programs to achieve high performance. Based on a linear inequalities mathematical model a formal specification of an optimized data decomposing algorithm and its implementation in C++ are presented. The algorithm enhances data locality and minimizes communication. Experimental results indicate that the algorithm improves the performance of parallelized programs significantly.