Enabling MPI interoperability through flexible communication endpoints

James Dinan, Pavan Balaji, David Goodell, Douglas Miller, Marc Snir, Rajeev Thakur

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

Abstract

The current MPI model defines a one-to-one relationship between MPI processes and MPI ranks. This model captures many use cases effectively, such as one MPI process per core and one MPI process per node. However, this semantic has limited interoperability between MPI and other programming models that use threads within a node. In this paper, we describe an extension to MPI that introduces communication endpoints as a means to relax the one-to-one relationship between processes and threads. Endpoints enable a greater degree interoperability between MPI and other programming models, and we illustrate their potential for additional performance and computation management benefits through the decoupling of ranks from processes.

Original languageEnglish (US)
Title of host publicationProceedings of the 20th European MPI Users' Group Meeting, EuroMPI 2013
PublisherAssociation for Computing Machinery
Pages13-18
Number of pages6
ISBN (Print)9788461651337
DOIs
StatePublished - 2013
Externally publishedYes
Event20th European MPI Users' Group Meeting, EuroMPI 2013 - Madrid, Spain
Duration: Sep 15 2013Sep 18 2013

Publication series

NameACM International Conference Proceeding Series

Other

Other20th European MPI Users' Group Meeting, EuroMPI 2013
Country/TerritorySpain
CityMadrid
Period9/15/139/18/13

Keywords

  • Endpoints
  • Hybrid parallel programming
  • Interoperability
  • MPI

ASJC Scopus subject areas

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

Fingerprint

Dive into the research topics of 'Enabling MPI interoperability through flexible communication endpoints'. Together they form a unique fingerprint.

Cite this