MPI Overlap: Benchmark and Analysis

Abstract : In HPC applications, one of the major overhead compared to sequentiel code, is communication cost. Application programmers often amortize this cost by overlapping communications with computation. To do so, they post a non-blocking MPI request, perform computation, and wait for communication completion, assuming MPI communication will progress in background. In this paper, we propose to measure what really happens when trying to overlap non-blocking point-to-point communications with computation. We explain how background progression works, we describe relevant test cases, we identify challenges for a benchmark, then we propose a benchmark suite to measure how much overlap happen in various cases. We exhibit overlap benchmark results on a wide panel of MPI libraries and hardware platforms. Finally, we classify, analyze, and explain the results using low-level traces to reveal the internal behavior of the MPI library.
Type de document :
Communication dans un congrès
International Conference on Parallel Processing, Aug 2016, Philadelphia, United States. 2016, 45th International Conference on Parallel Processing
Liste complète des métadonnées

Littérature citée [19 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01324179
Contributeur : <>
Soumis le : mardi 31 mai 2016 - 16:51:46
Dernière modification le : lundi 18 septembre 2017 - 09:52:05
Document(s) archivé(s) le : jeudi 1 septembre 2016 - 11:38:27

Fichier

final.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01324179, version 1
  • Mot de passe : jmioo2p

Citation

Alexandre Denis, François Trahay. MPI Overlap: Benchmark and Analysis. International Conference on Parallel Processing, Aug 2016, Philadelphia, United States. 2016, 45th International Conference on Parallel Processing. 〈hal-01324179〉

Partager

Métriques

Consultations de la notice

332

Téléchargements de fichiers

496