MPI+OpenMP Implementation and Results Analysis of Matrix Multiplication Based on Rowwise and Columnwise Block-Striped Decomposition of the Matrices

Author

He, Lidong ; Shen, Weichang ; Li, Yongjin ; Shi, Anlei ; Zhao, Dong

Author_Institution

Sch. of Inf. Sci. & Technol., Northwest Univ. NWU, Xi´´an, China

Volume

2

fYear

2010

fDate

28-31 May 2010

Firstpage

304

Lastpage

307

Abstract

This paper outlines the MPI+OpenMP programming model, and implements the matrix multiplication based on rowwise and columnwise block-striped decomposition of the matrices with MPI+OpenMP programming model in the multi-core cluster system. Experimental results show that the running time of the parallel algorithm is reduced significantly. By further analyzing the running results, the running time is reduced obviously once again.

Keywords

matrix decomposition; matrix multiplication; message passing; parallel algorithms; parallel programming; MPI; OpenMP implementation; columnwise block striped matrix decomposition; matrix multiplication; message passing interface; multi-core cluster system; parallel algorithm; rowwise block striped matrix decomposition; Central Processing Unit; Computer networks; Concurrent computing; Matrix decomposition; Parallel processing; Parallel programming; hybrid MPI/OpenMP programming model; matrix multiplication based on rowwise and columnwise block-striped decomposition of the matrices; multi-core cluster;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Science and Optimization (CSO), 2010 Third International Joint Conference on

Conference_Location

Huangshan

Print_ISBN

978-1-4244-6812-6

Electronic_ISBN

978-1-4244-6813-3

Type

conf

DOI

10.1109/CSO.2010.123

Filename

5533007