New adaptive GMRES(m) method with choosing suitable restart cycle m

Kentaro Moriya, Takashi Nodera

研究成果: Conference contribution

1 被引用数 (Scopus)

抄録

GMRES method is one of the major iterative algorithms for solving large and sparse linear systems of equations. However, it is difficult to implement GMRES algorithm because its storatege and computation cost are so exceeded. Therefore, GMRES(m) algorithm is often used. In this paper, we propose a new variant of GMRES(m) algorithm. Our algorithm chooses the restart cycle m based both on the convergence test of residual norm and on the distribution of zeros of residual polynomial of GMRES(m) algorithm. From the numerical examples on Compaq Beowulf, we also show the effectiveness of our proposed algorithm.

本文言語English
ホスト出版物のタイトルLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
編集者Roman Wyrzykowski, Jack Dongarra, Marcin Paprzycki, Jerzy Wasniewski
出版社Springer Verlag
ページ1105-1113
ページ数9
ISBN(印刷版)3540219463, 9783540219460
DOI
出版ステータスPublished - 2004 1 1
イベント5th International Conference on Parallel Processing and Applied Mathematics, PPAM 2003 - Czestochowa, Poland
継続期間: 2003 9 72003 9 10

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
3019
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Other

Other5th International Conference on Parallel Processing and Applied Mathematics, PPAM 2003
CountryPoland
CityCzestochowa
Period03/9/703/9/10

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

フィンガープリント 「New adaptive GMRES(m) method with choosing suitable restart cycle m」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル