Using node information to implement MPI cartesian topologies

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

Abstract

The MPI API provides support for Cartesian process topologies, including the option to reorder the processes to achieve better communication performance. But MPI implementations rarely provide anything useful for the reorder option, typically ignoring it. One argument made is that modern interconnects are fast enough that applications are less sensitive to the exact layout of processes onto the system. However, intranode communication performance is much greater than internode communication performance. In this paper, we show a simple approach that takes into account only information about which MPI processes are on the same node to provide a fast and effective implementation of the MPI Cartesian topology. While not optimal, this approach provides a significant improvement over all tested MPI implementations and provides an implementation that may be used as the default in any MPI implementation of MPI_Cart_create.

Original languageEnglish (US)
Title of host publicationEuroMPI 2018 - Proceedings of the 25th European MPI Users' Group Meeting
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450364928
DOIs
StatePublished - Sep 23 2018
Event25th European MPI Users' Group Meeting, EuroMPI 2018 - Barcelona, Spain
Duration: Sep 23 2018Sep 26 2018

Publication series

NameACM International Conference Proceeding Series

Other

Other25th European MPI Users' Group Meeting, EuroMPI 2018
CountrySpain
CityBarcelona
Period9/23/189/26/18

Keywords

  • Cartesian process topology
  • MPI
  • Message passing
  • Process topology

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'Using node information to implement MPI cartesian topologies'. Together they form a unique fingerprint.

Cite this