Load balancing for regular meshes on SMPs with MPI

Vivek Kale, William Gropp

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Domain decomposition for regular meshes on parallel computers has traditionally been performed by attempting to exactly partition the work among the available processors (now cores). However, these strategies often do not consider the inherent system noise which can hinder MPI application scalability to emerging peta-scale machines with 10000+ nodes. In this work, we suggest a solution that uses a tunable hybrid static/dynamic scheduling strategy that can be incorporated into current MPI implementations of mesh codes. By applying this strategy to a 3D jacobi algorithm, we achieve performance gains of at least 16% for 64 SMP nodes.

Original languageEnglish (US)
Title of host publicationRecent Advances in the Message Passing Interface - 17th European MPI Users' Group Meeting, EuroMPI 2010, Proceedings
Pages229-238
Number of pages10
DOIs
StatePublished - 2010
Event17th European MPI Users' Group Meeting, EuroMPI 2010 - Stuttgart, Germany
Duration: Sep 12 2010Sep 15 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6305 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other17th European MPI Users' Group Meeting, EuroMPI 2010
Country/TerritoryGermany
CityStuttgart
Period9/12/109/15/10

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Load balancing for regular meshes on SMPs with MPI'. Together they form a unique fingerprint.

Cite this