Umeå University's logo

umu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Evaluation of the Tunability of a New NUMA-Aware Hessenberg Reduction Algorithm
Umeå University, Faculty of Science and Technology, Department of Computing Science.
Umeå University, Faculty of Science and Technology, Department of Computing Science.ORCID iD: 0000-0002-4675-7434
Umeå University, Faculty of Science and Technology, Department of Computing Science.
2016 (English)Report (Other academic)
Abstract [en]

The reduction of a general dense and square matrix to Hessenberg form is a well known first step in many standard eigenvalue solvers. Although parallel algorithms exist, the Hessenberg reduction is still one of the bottlenecks in state-of-the-art software for the distributed QR algorithm. We propose a new NUMA-aware algorithm that fits the context of the QR algorithm and evaluate the tunability of its algorithmic parameters. The proposed algorithm can be faster than LAPACK and ScaLAPACK for small problem sizes. In addition, evaluating the algorithmic parameters shows that there is potential for auto-tuning some of the parameters.

Place, publisher, year, edition, pages
Umeå University , 2016. , p. 26
Series
Report / UMINF, ISSN 0348-0542 ; 16.21
Keywords [en]
Hessenberg reduction, parallel cache assignment, NUMA-aware algorithm, shared-memory algorithm, tunable parameters, off-line tuning
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:umu:diva-152576OAI: oai:DiVA.org:umu-152576DiVA, id: diva2:1255666
Available from: 2018-10-14 Created: 2018-10-14 Last updated: 2020-07-09Bibliographically approved

Open Access in DiVA

fulltext(650 kB)85 downloads
File information
File name FULLTEXT01.pdfFile size 650 kBChecksum SHA-512
8bb6a835160e0456d1b182f8c35a216f8f7b298b8dbf477e2114723eea299943345dedfac5458bc658e6d094a9d894c6202dbcd55834fc08d7719e5497479984
Type fulltextMimetype application/pdf

Other links

URL

Authority records

Eljammaly, MahmoudKarlsson, LarsKågström, Bo

Search in DiVA

By author/editor
Eljammaly, MahmoudKarlsson, LarsKågström, Bo
By organisation
Department of Computing Science
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 85 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 830 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf