Communication characterization and optimization of applications using topology-aware task mapping on large supercomputers

Sarat Sreepathi, Ed D'Azevedo, Bobby Philip, Patrick Worley

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

8 Scopus citations

Abstract

On large supercomputers, the job scheduling systems may assign a non-contiguous node allocation for user applications depending on available resources. With parallel applications using MPI (Message Passing Interface), the default process ordering does not take into account the actual physical node layout available to the application. This contributes to non-locality in terms of physical network topology and impacts communication performance of the application. In order to mitigate such performance penalties, this work describes techniques to identify suitable task mapping that takes the layout of the allocated nodes as well as the application's communication behavior into account. During the first phase of this research, we instrumented and collected performance data to characterize communication behavior of critical US DOE (United States - Department of Energy) applications using an augmented version of the mpiP tool. Subsequently, we developed several reordering methods (spectral bisection, neighbor join tree etc.) to combine node layout and application communication data for optimized task placement. We developed a tool called mpiAproxy to facilitate detailed evaluation of the various reordering algorithms without requiring full application executions. This work presents a comprehensive performance evaluation (14,000 experiments) of the various task mapping techniques in lowering communication costs on Titan, the leadership class supercomputer at Oak Ridge National Laboratory.

Original languageEnglish
Title of host publicationICPE 2016 - Proceedings of the 7th ACM/SPEC International Conference on Performance Engineering
PublisherAssociation for Computing Machinery, Inc
Pages225-236
Number of pages12
ISBN (Electronic)9781450340809
DOIs
StatePublished - Mar 12 2016
Event7th ACM/SPEC International Conference on Performance Engineering, ICPE 2016 - Delft, Netherlands
Duration: Mar 12 2016Mar 16 2016

Publication series

NameICPE 2016 - Proceedings of the 7th ACM/SPEC International Conference on Performance Engineering

Conference

Conference7th ACM/SPEC International Conference on Performance Engineering, ICPE 2016
Country/TerritoryNetherlands
CityDelft
Period03/12/1603/16/16

Keywords

  • Communication characterization
  • Reordering algorithms
  • Topology-aware optimization

Fingerprint

Dive into the research topics of 'Communication characterization and optimization of applications using topology-aware task mapping on large supercomputers'. Together they form a unique fingerprint.

Cite this