Fast solution of large N × N matrix equations in an MIMD-SIMD Hybrid System

Leo Chin Sim; C. G. (C. Graham) Leedham; Chin Jian Leo; Heiko Schröder

doi:10.1016/j.parco.2003.05.011

Fast solution of large N × N matrix equations in an MIMD-SIMD Hybrid System

Leo Chin Sim, C. G. (C. Graham) Leedham, Chin Jian Leo, Heiko Schröder

Research output: Contribution to journal › Article

3 Citations (Scopus)

Abstract

In this paper, we propose a new high-speed computation algorithm for solving a large N × N matrix system using the MIMD-SIMD Hybrid System. The MIMD-SIMD Hybrid System (also denoted as Hybrid System in this paper) is a new parallel architecture consisting of a combination of Cluster of Workstations (COWs) and SIMD systems working concurrently to produce an optimal parallel computation. We first introduce our prototype SIMD system and our Hybrid System setup before presenting how it can be implemented to find the unknowns in a large N × N linear matrix equation system using the Gauss-LU algorithm. This algorithm basically performs the 'Divide and Conquer' approach by breaking down the large N × N matrix system into a manageable 32 × 32 matrix for fast computation.

Original language	English
Pages (from-to)	1669-1684
Number of pages	16
Journal	Parallel Computing
Volume	29
Issue number	11-12 SPEC.ISS.
DOIs	https://doi.org/10.1016/j.parco.2003.05.011
Publication status	Published - 2003

Keywords

Computer algorithms
Matrices
Parallel processing (Electronic computers)
Speedup
Hybrid System
Cluster
Gauss-LU
Parallelization

Access to Document

10.1016/j.parco.2003.05.011

Cite this

@article{6461c1d05cf747acaa352fd28660516d,

title = "Fast solution of large N × N matrix equations in an MIMD-SIMD Hybrid System",

abstract = "In this paper, we propose a new high-speed computation algorithm for solving a large N × N matrix system using the MIMD-SIMD Hybrid System. The MIMD-SIMD Hybrid System (also denoted as Hybrid System in this paper) is a new parallel architecture consisting of a combination of Cluster of Workstations (COWs) and SIMD systems working concurrently to produce an optimal parallel computation. We first introduce our prototype SIMD system and our Hybrid System setup before presenting how it can be implemented to find the unknowns in a large N × N linear matrix equation system using the Gauss-LU algorithm. This algorithm basically performs the 'Divide and Conquer' approach by breaking down the large N × N matrix system into a manageable 32 × 32 matrix for fast computation.",

keywords = "Computer algorithms, Matrices, Parallel processing (Electronic computers), Speedup, Hybrid System, Cluster, Gauss-LU, Parallelization",

author = "{Chin Sim}, Leo and Leedham, {C. G. (C. Graham)} and Leo, {Chin Jian} and Heiko Schr{\"o}der",

year = "2003",

doi = "10.1016/j.parco.2003.05.011",

language = "English",

volume = "29",

pages = "1669--1684",

journal = "Parallel Computing",

issn = "0167-8191",

publisher = "Elsevier North-Holland",

number = "11-12 SPEC.ISS.",

}

TY - JOUR

T1 - Fast solution of large N × N matrix equations in an MIMD-SIMD Hybrid System

AU - Chin Sim, Leo

AU - Leedham, C. G. (C. Graham)

AU - Leo, Chin Jian

AU - Schröder, Heiko

PY - 2003

Y1 - 2003

N2 - In this paper, we propose a new high-speed computation algorithm for solving a large N × N matrix system using the MIMD-SIMD Hybrid System. The MIMD-SIMD Hybrid System (also denoted as Hybrid System in this paper) is a new parallel architecture consisting of a combination of Cluster of Workstations (COWs) and SIMD systems working concurrently to produce an optimal parallel computation. We first introduce our prototype SIMD system and our Hybrid System setup before presenting how it can be implemented to find the unknowns in a large N × N linear matrix equation system using the Gauss-LU algorithm. This algorithm basically performs the 'Divide and Conquer' approach by breaking down the large N × N matrix system into a manageable 32 × 32 matrix for fast computation.

AB - In this paper, we propose a new high-speed computation algorithm for solving a large N × N matrix system using the MIMD-SIMD Hybrid System. The MIMD-SIMD Hybrid System (also denoted as Hybrid System in this paper) is a new parallel architecture consisting of a combination of Cluster of Workstations (COWs) and SIMD systems working concurrently to produce an optimal parallel computation. We first introduce our prototype SIMD system and our Hybrid System setup before presenting how it can be implemented to find the unknowns in a large N × N linear matrix equation system using the Gauss-LU algorithm. This algorithm basically performs the 'Divide and Conquer' approach by breaking down the large N × N matrix system into a manageable 32 × 32 matrix for fast computation.

KW - Computer algorithms

KW - Matrices

KW - Parallel processing (Electronic computers)

KW - Speedup

KW - Hybrid System

KW - Cluster

KW - Gauss-LU

KW - Parallelization

UR - http://handle.uws.edu.au:8081/1959.7/10554

UR - http://www.scopus.com/inward/record.url?scp=0242574588&partnerID=8YFLogxK

U2 - 10.1016/j.parco.2003.05.011

DO - 10.1016/j.parco.2003.05.011

M3 - Article

SN - 0167-8191

VL - 29

SP - 1669

EP - 1684

JO - Parallel Computing

JF - Parallel Computing

IS - 11-12 SPEC.ISS.

ER -

Fast solution of large N × N matrix equations in an MIMD-SIMD Hybrid System

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this