Measuring and Interpreting Dependent Task-Based Applications Performances

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

Abstract

Breaking down the parallel time into work, idleness, and overheads is crucial for assessing the performance of HPC applications, but difficult to measure in asynchronous dependent tasking runtime systems. No existing tools allow its measurement portably and accurately. This paper introduces POT: a tool-suite for dependent task-based applications performance measurement. We focus on its low-disturbance methodology consisting of task modeling, discrete-event tracing, and post-mortem simulation-based analysis. It supports the OMPT standard OpenMP specifications.

Original languageEnglish
Title of host publicationParallel Processing and Applied Mathematics - 15th International Conference, PPAM 2024, Revised Selected Papers
EditorsRoman Wyrzykowski, Jack Dongarra, Ewa Deelman, Konrad Karczewski
PublisherSpringer Science and Business Media Deutschland GmbH
Pages227-241
Number of pages15
ISBN (Print)9783031856969
DOIs
Publication statusPublished - 1 Jan 2025
Externally publishedYes
Event15th International Conference on Parallel Processing and Applied Mathematics, PPAM 2024 - Ostrava, Czech Republic
Duration: 8 Sept 202411 Sept 2024

Publication series

NameLecture Notes in Computer Science
Volume15579
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference15th International Conference on Parallel Processing and Applied Mathematics, PPAM 2024
Country/TerritoryCzech Republic
CityOstrava
Period8/09/2411/09/24

Keywords

  • OpenMP
  • Performances
  • Tasks
  • Time Breakdown

Fingerprint

Dive into the research topics of 'Measuring and Interpreting Dependent Task-Based Applications Performances'. Together they form a unique fingerprint.

Cite this