Beta
33937

Development an Analytical Performance Models for Matrix Multiplication on Distributing Systems

Article

Last updated: 24 Dec 2024

Subjects

-

Tags

-

Abstract

In this paper, we suggest a mechanism for implementing a distributed application using RMI based on JAVA threads. The application is parallel matrices multiplication depending on distributed the products block of rows and columns on different machines. One server and seven clients are run to find the product of matrix multiplication. The server distributes the determine blocks of rows and columns on the registered clients. The clients return their product blocks to a server, which
calculate the final product of matrix multiplication. Applications of this type will allow loaded servers to transfer part of the load to clients to exploit the computing power available at client side. Experimental result shows that the speed up ratio is equals 9 or the computation time of matrix multiplication with size of 2048 X 2048 is reduced by 89 % by using 7-client.

DOI

10.21608/ijci.2009.33937

Keywords

Distributed Systems, Performance prediction model, Matrix multiplication

Authors

First Name

Arabi

Last Name

Keshk

MiddleName

-

Affiliation

Faculty of Computer and Information Menoufia University

Email

arabikeshk@yahoo.com

City

-

Orcid

-

Volume

2

Article Issue

1

Related Issue

5671

Issue Date

2009-06-01

Receive Date

2009-01-04

Publish Date

2009-06-01

Page Start

41

Page End

52

Print ISSN

1687-7853

Online ISSN

2735-3257

Link

https://ijci.journals.ekb.eg/article_33937.html

Detail API

https://ijci.journals.ekb.eg/service?article_code=33937

Order

4

Type

Original Article

Type Code

877

Publication Type

Journal

Publication Title

IJCI. International Journal of Computers and Information

Publication Link

https://ijci.journals.ekb.eg/

MainTitle

Development an Analytical Performance Models for Matrix Multiplication on Distributing Systems

Details

Type

Article

Created At

22 Jan 2023