Design, tuning and evaluation of parallel multilevel ILU preconditioners

José I. Aliaga, Matthias Bollhöfer, Alberto F. Martín, Enrique S. Quintana-Ortí

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

5 Citations (Scopus)

Abstract

In this paper, we present a parallel multilevel ILU preconditioner implemented with OpenMP. We employ METIS partitioning algorithms to decompose the computation into concurrent tasks, which are then scheduled to threads. Concretely, we combine decompositions which obtain significantly more tasks than processors, and the use of dynamic scheduling strategies in order to reduce the thread's idle time, which it is shown to be the main source of overhead in our parallel algorithm. Experimental results on a shared-memory platform consisting of 16 processors report remarkable performance for our approach.

Original languageEnglish
Title of host publicationHigh Performance Computing for Computational Science - VECPAR 2008 - 8th International Conference, Revised Selected Papers
Pages314-327
Number of pages14
DOIs
Publication statusPublished - 1 Dec 2008
Externally publishedYes
EventInternational Conference on High Performance Computing for Computational Science, 2008 - Toulouse, France
Duration: 24 Jun 200827 Jun 2008
Conference number: 8th

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5336 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceInternational Conference on High Performance Computing for Computational Science, 2008
Abbreviated titleVECPAR 2008
CountryFrance
CityToulouse
Period24/06/0827/06/08

Keywords

  • Incomplete LU factorization
  • OpenMP
  • Parallel algorithm
  • Shared-memory multiprocessor
  • Sparse linear system

Cite this