Algorithmic scheme for hybrid computing with CPU, Xeon-Phi/MIC and GPU devices on a single machine

Sylvain Contassot-Vivier 1 Stéphane Vialle 2, 3
1 ALGORILLE - Algorithms for the Grid
Inria Nancy - Grand Est, LORIA - NSS - Department of Networks, Systems and Services
2 UMI2958 - Georgia Tech - CNRS [Metz]
3 SUPELEC-Campus Metz
Abstract : In this paper, we address the problem of the efficient parallel exploitation of different types of computing devices inside a single machine, to solve a scientific problem. As a first step, we apply our scheme to the Jacobi relaxation. Despite its simplicity, it is a good example of iterative process for scientific simulation. Then, we evaluate and analyze the performance of our parallel implementation on two configurations of hybrid machine.
Keywords : offload computing heterogeneous computing Accelerator Xeon-Phi/MIC GPU hybrid computing
Type de document :
Communication dans un congrès
International Conference on Parallel Computing 2015 (ParCo 2015), Sep 2015, Edinburgh, United Kingdom. pp.1-10, 2015
Liste complète des métadonnées
https://hal-centralesupelec.archives-ouvertes.fr/hal-01258499
Contributeur : Sébastien Van Luchene <>
Soumis le : mardi 19 janvier 2016 - 10:59:33
Dernière modification le : jeudi 5 avril 2018 - 12:30:11

Identifiants

  • HAL Id : hal-01258499, version 1

Collections

UNIV-FCOMTE | INRIA | UMI-GTL | UNIV-LORRAINE | CENTRALESUPELEC | IDMAD | UMI-COMPUTERSCIENCE | LORIA-NSS | LORIA | UNIV-PARIS-SACLAY

Citation

Sylvain Contassot-Vivier, Stéphane Vialle. Algorithmic scheme for hybrid computing with CPU, Xeon-Phi/MIC and GPU devices on a single machine. International Conference on Parallel Computing 2015 (ParCo 2015), Sep 2015, Edinburgh, United Kingdom. pp.1-10, 2015. 〈hal-01258499〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

519

Contact

support.ccsd.cnrs.fr
hal.support@ccsd.cnrs.fr
Logo CCSD