Show simple item record

dc.contributor.authorHaidar, Azzam
dc.contributor.authorLtaief, Hatem
dc.contributor.authorDongarra, Jack
dc.date.accessioned2015-05-25T08:28:46Z
dc.date.available2015-05-25T08:28:46Z
dc.date.issued2012-01
dc.identifier.citationToward a High Performance Tile Divide and Conquer Algorithm for the Dense Symmetric Eigenvalue Problem 2012, 34 (6):C249 SIAM Journal on Scientific Computing
dc.identifier.issn1064-8275
dc.identifier.issn1095-7197
dc.identifier.doi10.1137/110823699
dc.identifier.urihttp://hdl.handle.net/10754/555650
dc.description.abstractClassical solvers for the dense symmetric eigenvalue problem suffer from the first step, which involves a reduction to tridiagonal form that is dominated by the cost of accessing memory during the panel factorization. The solution is to reduce the matrix to a banded form, which then requires the eigenvalues of the banded matrix to be computed. The standard divide and conquer algorithm can be modified for this purpose. The paper combines this insight with tile algorithms that can be scheduled via a dynamic runtime system to multicore architectures. A detailed analysis of performance and accuracy is included. Performance improvements of 14-fold and 4-fold speedups are reported relative to LAPACK and Intel's Math Kernel Library.
dc.publisherSociety for Industrial & Applied Mathematics (SIAM)
dc.relation.urlhttp://epubs.siam.org/doi/abs/10.1137/110823699
dc.rightsArchived with thanks to SIAM Journal on Scientific Computing
dc.subjectdivide and conquer
dc.subjectsymmetric eigenvalue solver
dc.subjecttile algorithms
dc.subjectdynamic scheduling
dc.titleToward a High Performance Tile Divide and Conquer Algorithm for the Dense Symmetric Eigenvalue Problem
dc.typeArticle
dc.contributor.departmentKAUST Supercomputing Laboratory (KSL)
dc.identifier.journalSIAM Journal on Scientific Computing
dc.eprint.versionPublisher's Version/PDF
dc.contributor.institutionDepartment of Electrical Engineering and Computer Science, University of Tennessee, Knoxville, TN
dc.contributor.institutionComputer Science and Mathematics Division, Oak Ridge National Laboratory, Oak Ridge, TN
dc.contributor.institutionSchool of Mathematics and School of Computer Science, University of Manchester, Manchester, UK
kaust.personLtaief, Hatem
refterms.dateFOA2018-06-14T07:55:40Z


Files in this item

Thumbnail
Name:
110823699.pdf
Size:
688.6Kb
Format:
PDF
Description:
Main article

This item appears in the following Collection(s)

Show simple item record