QMPI: A next generation MPI profiling interface for modern HPC platforms

Bengisu Elis, Dai Yang, Martin Schulz

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

1 Scopus citations

Abstract

As we approach exascale and start planning for beyond, the rising complexity of systems and applications demands new monitoring, analysis, and optimization approaches. This requires close coordination with the parallel programming system used, which for HPC in most cases includes MPI, the Message Passing Interface. While MPI provides comprehensive tool support in the form of the MPI Profiling interface, PMPI, which has inspired a generation of tools, it is not sufficient for the new arising challenges. In particular, it does not support modern software design principles nor the composition of multiple monitoring solutions from multiple agents or sources. We approach these gaps and present QMPI, as a possible successor to PMPI. In this paper, we present the use cases and requirements that drive its development, offer a prototype design and implementation, and demonstrate its effectiveness and low overhead.

Original languageEnglish
Title of host publicationProceedings of the 26th European MPI Users'' Group Meeting, EuroMPI 2019
EditorsTorsten Hoefler, Jesper Larsson Traff
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450371759
DOIs
StatePublished - 11 Sep 2019
Event26th European MPI Users'' Group Meeting, EuroMPI 2019 - Zurich, Switzerland
Duration: 11 Sep 201913 Sep 2019

Publication series

NameACM International Conference Proceeding Series

Conference

Conference26th European MPI Users'' Group Meeting, EuroMPI 2019
Country/TerritorySwitzerland
CityZurich
Period11/09/1913/09/19

Keywords

  • Dynamic tool-chain
  • MPI Profiling Interface
  • MPI tools
  • PMPI

Fingerprint

Dive into the research topics of 'QMPI: A next generation MPI profiling interface for modern HPC platforms'. Together they form a unique fingerprint.

Cite this