三項対角近似因子分解による前処理共役勾配法: スーパコン向き非対称連立一次方程式解法

Material type: 文書・図像類

Author: 土肥, 俊ほか

Publisher: 航空宇宙技術研究所

Publication date: 1986-12

Material Format: Digital

Capacity, size, etc.: -

NDC: -

View All

Notes on use

Note (General)：: In this paper, we propose the Tridiagonal Approximate Factorization (TF) method as a preconditioning algorithm suitable to vector and parallel compute...

Search by Bookstore

Holdings of Libraries in Japan

This page shows libraries in Japan other than the National Diet Library that hold the material.

List of Cooperating Institutions and Databases

Please contact your local library for information on how to use materials or whether it is possible to request materials from the holding libraries.

other

JAXA Repository
Digital
You can check the holdings of institutions and databases with which Institutional Repositories DataBase(IRDB)(Institutional Repository) is linked at the site of Institutional Repositories DataBase(IRDB)(Institutional Repository).
JAXA Repository

Search by Bookstore

Bibliographic Record

You can check the details of this material, its authority (keywords that refer to materials on the same subject, author's name, etc.), etc.

Digital

Material Type: 文書・図像類
Title: 三項対角近似因子分解による前処理共役勾配法: スーパコン向き非対称連立一次方程式解法
Author/Editor: 土肥, 俊
原田, 紀夫
Doi, Shun
Harada, Norio
Author Heading: 土肥, 俊
原田, 紀夫
Doi, Shun
Harada, Norio
Publication, Distribution, etc.: 航空宇宙技術研究所
Publication Date: 1986-12
Publication Date (W3CDTF): 1986-12
Alternative Title: Tridiagonal Approximate Factorization Method : A Preconditioning Technique for Solving Nonsymmetric Linear Systems Suitable to Supercomputers
Periodical title: 航空宇宙技術研究所特別資料 = Special Publication of National Aerospace Laboratory SP-7
No. or year of volume/issue: 7
Volume: 7
Pages: 143-150
ISSN (Periodical Title): ISSN : 0289-260X
Text Language Code: jpn
Target Audience: 一般
Note (General): In this paper, we propose the Tridiagonal Approximate Factorization (TF) method as a preconditioning algorithm suitable to vector and parallel computers for solving large sparce nonsymmetric systems of linear equations. In this method, the preconditioning matrix is constructed as a product of component tridiagonal matrices, each of which consists of matrix elements corresponding to a spatial axis of the original P. D. E. problem. The inversion process of this preconditioning matrix is reduced to a series of simple forward-and-backward substitutions, which have inherent parallelism and hence can be easily vectorized. Comparison is made between the TFBCG algorithm, which applies TF preconditioning to the Bi-Conjugate Gradient algorithm, and the ILUBCG/MILUBCG algorithms, the most powerful and popular methods on conventional scaler computers. Numerical tests on the SX-2 supercomputer show that the TFBCG program is 3 to 9 times faster than the conventional (M) ILUBCG programs, and also 1 to 4 times faster than the sophisticated (M) ILUBCG programs utilizing the list-vector technique.
資料番号: NALSP0007018
レポート番号: NAL SP-7
Access Restrictions: 限定公開
Periodical Title (NCID): AN10097345
https://ci.nii.ac.jp/ncid/AN10097345
Data Provider (Database): 国立情報学研究所 : 学術機関リポジトリデータベース（IRDB）（機関リポジトリ）
https://irdb.nii.ac.jp
Original Data Provider (Database): 宇宙航空研究開発機構 : 宇宙航空研究開発機構リポジトリ