Singer, Sanja and Singer, Saša and Novaković, Vedran and Davidović, Davor and Bokulić, Krešimir and Ušćumlić, Aleksandar (2012) Three-level parallel J-Jacobi algorithms for Hermitian matrices. = Three-level parallel J-Jacobi algorithms for Hermitian matrices. Applied Mathematics and Computation, 218 (9). pp. 5704-5725. ISSN 0096-3003. Vrsta rada: ["eprint_fieldopt_article_type_article" not defined]. Kvartili JCR: Q1 (2012). .
|
Text
Dummy article.pdf - Draft Version Jezik dokumenta:English Download (79kB) | Preview |
Abstract
The paper describes several efficient parallel implementations of the one-sided hyperbolic Jacobi-type algorithm for computing eigenvalues and eigenvectors of Hermitian matrices. By appropriate blocking of the algorithms an almost ideal load balancing between all available processors/cores is obtained. A similar blocking technique can be used to exploit local cache memory of each processor to further speed up the process. Due to diversity of modern computer architectures, each of the algorithms described here may be the method of choice for a particular hardware and a given matrix size. All proposed block algorithms compute the eigenvalues with relative accuracy similar to the original non-blocked Jacobi algorithm. © 2011 Elsevier Inc. All rights reserved.
Item Type: | Article (["eprint_fieldopt_article_type_article" not defined]) |
---|---|
Keywords (Croatian): | Block strategies; Blocking; Eigenvalues; Hermitian matrices; Parallelizations; Algorithms; Cache memory; Computer architecture; Computer hardware; Efficiency; Eigenvalues and eigenfunctions; Jacobian matrices |
Subjects: | NATURAL SCIENCES > Mathematics TECHNICAL SCIENCE |
Divisions: | 1500 Chair of Mathematics |
Indexed in Web of Science: | Yes |
Indexed in Current Contents: | Yes |
Citations JCR: | 3 (27.4.2015.) |
Quartiles: | Q1 (2012) |
Citations SCOPUS: | 2 (27.4.2015.) |
Date Deposited: | 27 Apr 2015 09:20 |
Last Modified: | 23 May 2017 11:20 |
URI: | http://repozitorij.fsb.hr/id/eprint/3975 |
Actions (login required)
![]() |
View Item |