Generating efficient parallel code for successive over-relaxation

被引：0

作者：

Tang, PY ^{[1
]}

机构：

[1] Univ So Queensland, Dept Math & Comp, Toowoomba, Qld 4350, Australia

来源：

ICA(3)PP 97 - 1997 3RD INTERNATIONAL CONFERENCE ON ALGORITHMS AND ARCHITECTURES FOR PARALLEL PROCESSING | 1997年

关键词：

D O I：

暂无

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

A complete suite of algorithms for parallelizing compilers to generate efficient SPMD code for SOR problems is presented. By applying unimodular transformation before loop tiling and parallelization, the number of messages per iteration per processor is reduced from 3(n) - 1 in the conventional parallel SOR algorithm to 2(n) - 1, where n is the dimensionality of the data set. To maintain the memory-scalability, a novel approach to use the local dynamic memory of parallel processors to implement the skewed data set is proposed.

引用

页码：503 / 510

页数：8