Collective communication on architectures that support simultaneous communication over multiple links

Ernie Chan, Robert Van De Geijn, William Gropp, Rajeev Thakur

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

Abstract

Traditional collective communication algorithms are designed with the assumption that a node can communicate with only one other node at a time. On new parallel architectures such as the IBM Blue Gene/L, a node can communicate with multiple nodes simultaneously. We have redesigned and reimplemented many of the MPI collective communication algorithms to take advantage of this ability to send simultaneously, including broadcast, reduce(-to-one), scatter, gather, allgather, reduce-scatter, and allreduce. We show that these new algorithms have lower expected costs than the previously known lower bounds based on old models of parallel computation. Results are included comparing their performance to the default implementations in IBM's MPI.

Original languageEnglish (US)
Title of host publicationProceedings of the 2006 ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP'06
PublisherAssociation for Computing Machinery
Pages2-11
Number of pages10
ISBN (Print)1595931899, 9781595931894
DOIs
StatePublished - 2006
Externally publishedYes
Event2006 ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP'06 - New York, NY, United States
Duration: Mar 29 2006Mar 31 2006

Publication series

NameProceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP
Volume2006

Other

Other2006 ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP'06
Country/TerritoryUnited States
CityNew York, NY
Period3/29/063/31/06

Keywords

  • Algorithms
  • Performance

ASJC Scopus subject areas

  • Software

Fingerprint

Dive into the research topics of 'Collective communication on architectures that support simultaneous communication over multiple links'. Together they form a unique fingerprint.

Cite this