Skip to main navigation Skip to search Skip to main content

Finding Hamiltonian paths in tournaments on clusters - A provably communication-efficient approach

  • SUNY Buffalo

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

3 Scopus citations

Abstract

This paper presents a general methodology for the communication efficient parallelization of graph algorithms using divide-and-conquer approach and shows that this class of problems can be solved in cluster environments with good communication efficiency. Specifically, the first practical parallel algorithm, based onEREW BSP model, for finding Hamiltonian paths in tournaments is presented. For all commercially available parallel computing environments, this algorithm uses only (3logp+1) communication supersteps, which is independent of the tournament size, and can reuse the existing linear-time algorithm in sequential setting. Experiments have been carried out on a Linux cluster of 32 Sun Ultra5 computers and SGI Origin 2000 with 32 R10000 processors, using MPI. The algorithm performance on Linux Cluster reaches 75% of the performance on SGI Origin 2000 when the tournament size is about one million.

Original languageEnglish
Title of host publicationProceedings of the 2001 ACM Symposium on Applied Computing, SAC 2001
PublisherAssociation for Computing Machinery
Pages549-553
Number of pages5
ISBN (Print)1581132875, 9781581132878
DOIs
StatePublished - Mar 1 2001
Event2001 ACM Symposium on Applied Computing, SAC 2001 - Las Vegas, United States
Duration: Mar 11 2001Mar 14 2001

Publication series

NameProceedings of the ACM Symposium on Applied Computing

Conference

Conference2001 ACM Symposium on Applied Computing, SAC 2001
Country/TerritoryUnited States
CityLas Vegas
Period03/11/0103/14/01

Keywords

  • BSP
  • Cluster computing
  • MPI
  • Tournament

Fingerprint

Dive into the research topics of 'Finding Hamiltonian paths in tournaments on clusters - A provably communication-efficient approach'. Together they form a unique fingerprint.

Cite this