No Access

PERFORMANCE STUDY OF LU FACTORIZATION WITH LOW COMMUNICATION OVERHEAD ON MULTIPROCESSORS

F. DESPREZ

The University of Tennessee, Computer Science Department, 107, Ayres Hall, Knoxville, TN 37996-1301, USA

On leave from LIP, CNRS URA 1398, ENS Lyon, 46 allée d'Italie, 69364 Lyon Cedex 07, France.

This work was supported by MRE grant No. 974, the CNRS-NSF grant No. 950.22/07 and the research program C3.

Search for more papers by this author

J. J. DONGARRA

The University of Tennessee, Computer Science Department, 107, Ayres Hall, Knoxville, TN 37996-1301, USA

This work was supported by MRE grant No. 974, the CNRS-NSF grant No. 950.22/07 and the research program C3.

This work was supported in part by the National Science Foundation under grant ASC-8715728, the National Science Foundation Science and Technology Center Cooperative Agreement CCR-8809615, the DARPA and ARO under contract DAAL03–91-C-0047.

Search for more papers by this author

, and

B. TOURANCHEAU

The University of Tennessee, Computer Science Department, 107, Ayres Hall, Knoxville, TN 37996-1301, USA

On leave from LIP, CNRS URA 1398, ENS Lyon, 46 allée d'Italie, 69364 Lyon Cedex 07, France.

This work was supported in part by CNRS-NSF grant 950.223/07, Archipel SA and MRE under grant 974, the National Science Foundation under grant ASC-8715728, the National Science Foundation Science and Technology Center Cooperative Agreement CCR-8809615, the DARPA and ARO under contract DAAL03-91-C-0047, PRC C3, and the DRET.

Search for more papers by this author

https://doi.org/10.1142/S012962649500014XCited by:8 (Source: Crossref)

Abstract

In this paper, we make efficient use of asynchronous communications on the LU decomposition algorithm with pivoting and a column-scattered data decomposition to derive precise computational complexities. We then compare these results with experiments on the Intel iPSC/860 and Paragon machines and show that very good performances can be obtained on a ring with asynchronous communications.

Keywords: