Advertisement

Simplifying communication induced by operations on block-distributed arrays

  • Andreas Eberhart
  • Jingke Li
Workshop 03 Automatic Parallelization and High Performance Compilers
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1123)

Abstract

This paper analyzes communication patterns induced by operations on arrays with arbitrary block-distributions. A simple method is introduced that transforms the resulting pattern into a collection of independent gathers and scatters, which can be implemented efficiently.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    M. Barnett, S. Gupta, D. G. Payne, L. Shuler, R. A. van de Geijn, and J. Watts. Building a high-performance collective communication library. In Proceedings of Supercomputing '94, pages 107–116, 1994.Google Scholar
  2. 2.
    A. Eberhart and J. Li. Optimizing communication for array operations on message-passing multicomputers. In Proceedings of the 1996 ISPANN, 1996.Google Scholar
  3. 3.
    G. Fox, M. Johnson, G. Lyzenga, S. Otto, J. Salmon, and D. Walker. Solving Problems on Concurrent Processors. Prentice Hall, 1988.Google Scholar
  4. 4.
    J. Li and M. Chen. Compiling communication-efficient programs for massively-parallel machines. IEEE Trans. on Parallel and Dist. Systems, 2(3), July 1991.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Andreas Eberhart
    • 1
  • Jingke Li
    • 1
  1. 1.Portland State UniversityPortlandUSA

Personalised recommendations