A parallel implementation of GESPP on a cluster of Silicon Graphics workstations

Author

Al-Towaiq, Mohammad ; Al-Aamri, Huda

Author_Institution

Dept. of Comput. Sci., Sultan Qaboos Univ., Muscat, Oman

fYear

2002

fDate

17-20 Dec. 2002

Firstpage

226

Lastpage

230

Abstract

In this paper, we develop and evaluate a parallel algorithm for GESPP on a cluster of Silicon Graphics workstations using PVM as a parallel programming environment. The proposed algorithm reduces the computation time to O(n³ / p) and the communication time to O(n) including scaling, searching for the pivot element, row interchanging, and pivot row and multipliers column broadcasting. The numerical experiments show that the proposed parallel algorithm offers a high speedup and efficiency for the test problems.

Keywords

computational complexity; matrix algebra; parallel algorithms; parallel programming; workstation clusters; GESPP; PVM; Silicon Graphics workstation cluster; communication time; computation time; multiplier column broadcasting; numerical experiments; parallel algorithm; parallel programming environment; pivot element; pivot row broadcasting; row interchanging; scaling; Application software; Clustering algorithms; Computer graphics; Concurrent computing; Equations; Load management; Matrix decomposition; Parallel algorithms; Silicon; Workstations;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel and Distributed Systems, 2002. Proceedings. Ninth International Conference on

ISSN

1521-9097

Print_ISBN

0-7695-1760-9

Type

conf

DOI

10.1109/ICPADS.2002.1183404

Filename

1183404