应用科学学报 ›› 1998, Vol. 16 ›› Issue (4): 469-478.

• 论文 • 上一篇    下一篇

一类新的广义同步与异步并行矩阵多分裂松弛算法

白中治1, 苏仰锋2   

  1. 1. 中国科学院计算数学与科学工程计算研究所;
    2. 复旦大学
  • 收稿日期:1995-11-04 修回日期:1996-02-04 出版日期:1998-12-31 发布日期:1998-12-31
  • 作者简介:白中治,Institute of Computational Mathematics and Scientific/Engineering Computing,Chinese Academy of Sciences,Beijing 100080

New Variants of the Generalized Synchronous and Asynchronous Parallel Matrix Multisplitting Relaxation Methods

BAI ZHONGZHI1, SU YANGFENG2   

  1. 1. State Key Laboratory of Scientific/Engineering Computing Institute of Computational Mathematics and Scientific/Engineering Computing Chinese Academy of Sciences, Beijing 100080;
    2. Department of Mathematics, Fudan Universit
  • Received:1995-11-04 Revised:1996-02-04 Online:1998-12-31 Published:1998-12-31

摘要: 对于并行求解大型稀疏线性代数方程组的同步与异步并行矩阵多分裂向前向后松弛算法,提出了分别适用于SIMD和MIMD多处理机系统的有效变型;并在通常条件下,建立了它们的收敛理论.

关键词: 矩阵多分裂, 收敛理论, 线性方程组, 同步与异步迭代, 松弛算法

Abstract: New variants of the existing synchronous and asynchronous parallel matrix multisplitting forward and backward relaxation methods for parallely solving the large sparse system of linear equations on the SIMD and MIMD multiprocessor systems are proposed, and their convergence theories are established under the ordinary conditions.

Key words: matrix multisplitting, synchronous and asynchronous iterations, system of linear equations, relaxed methods, convergence theory