Fast solution of large N × N matrix equations in an MIMD-SIMD hybrid system

Authors:
Leo Chin Sim;Graham Leedham;Leo Chin Jian;Heiko Schroder
Affiliations:
Centre for High Performance Embedded Systems, School of Computer Engineering, Nanyang Technological University (NTU), Blk N4, 2A-32, Nanyang Avenue, Singapore, Singapore;Centre for High Performance Embedded Systems, School of Computer Engineering, Nanyang Technological University (NTU), Blk N4, 2A-32, Nanyang Avenue, Singapore, Singapore;School of Engineering and Industrial Design, University of Western Sydney, Australia;School of Computer and IT, Royal Melbourne Institute of Technology, Australia
Venue:
Parallel Computing - Special issue: Parallel and distributed scientific and engineering computing
Year:
2003

Citing 11
Cited 1

Communication effect basic linear algebra computations on hypercube architectures

Journal of Parallel and Distributed Computing
Instruction systolic array—tradeoff between flexibility and speed

Computer Systems Science and Engineering
Parallel solution of triangular systems on distributed-memory multiprocessors

SIAM Journal on Scientific and Statistical Computing
The Spectral Decomposition of Nonsymmetric Matrices on Distributed Memory Parallel Computers

SIAM Journal on Scientific Computing
LAPACK Users' guide (third ed.)

LAPACK Users' guide (third ed.)
Matrix Multiplication on Heterogeneous Platforms

IEEE Transactions on Parallel and Distributed Systems
Dense linear algebra kernels on heterogeneous platforms: redistribution issues

Parallel Computing - Parallel matrix algorithms and applications
Morphological Hough Transform on the Instruction Systolic Array

Euro-Par '97 Proceedings of the Third International Euro-Par Conference on Parallel Processing
MIMD-SIMD hybrid system: towards a new low cost parallel system

Parallel Computing
LAPACK Working Note 41: Installation Guide for LAPACK

LAPACK Working Note 41: Installation Guide for LAPACK
Performance of Various Computers Using Standard Linear Equations Software

Performance of Various Computers Using Standard Linear Equations Software

Efficient biorthogonal Lanczos algorithm on message passing parallel computer

MTPP'10 Proceedings of the Second Russia-Taiwan conference on Methods and tools of parallel programming multicomputers

Quantified Score

Hi-index	0.00

Visualization

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.