Impact of asynchronism on GPU accelerated parallel iterative computations

International audience We study the impact of asynchronism on parallel iterative algorithms in the particular context of local clusters of workstations including GPUs. The application test is a classical PDE problem of advection-diffusion-reaction in 3D. We propose an asynchronous version of a previ...

Full description

Bibliographic Details
Main Authors: Contassot-Vivier, Sylvain, Jost, Thomas, Vialle, Stéphane
Other Authors: Algorithms for the Grid (ALGORILLE), INRIA Lorraine, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS), SUPELEC-Campus Metz, Ecole Supérieure d'Electricité - SUPELEC (FRANCE)
Format: Conference Object
Language:English
Published: HAL CCSD 2010
Subjects:
Online Access:https://hal-supelec.archives-ouvertes.fr/hal-00491976
https://hal-supelec.archives-ouvertes.fr/hal-00491976/document
https://hal-supelec.archives-ouvertes.fr/hal-00491976/file/Supelec620.pdf
id ftccsdartic:oai:HAL:hal-00491976v1
record_format openpolar
spelling ftccsdartic:oai:HAL:hal-00491976v1 2023-05-15T16:48:06+02:00 Impact of asynchronism on GPU accelerated parallel iterative computations Contassot-Vivier, Sylvain Jost, Thomas Vialle, Stéphane Algorithms for the Grid (ALGORILLE) INRIA Lorraine Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA) Institut National de Recherche en Informatique et en Automatique (Inria)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS) SUPELEC-Campus Metz Ecole Supérieure d'Electricité - SUPELEC (FRANCE) Reykjavik, Iceland 2010-06-06 https://hal-supelec.archives-ouvertes.fr/hal-00491976 https://hal-supelec.archives-ouvertes.fr/hal-00491976/document https://hal-supelec.archives-ouvertes.fr/hal-00491976/file/Supelec620.pdf en eng HAL CCSD hal-00491976 https://hal-supelec.archives-ouvertes.fr/hal-00491976 https://hal-supelec.archives-ouvertes.fr/hal-00491976/document https://hal-supelec.archives-ouvertes.fr/hal-00491976/file/Supelec620.pdf info:eu-repo/semantics/OpenAccess Proceedings of PARA 2010 conference: State of the Art in Scientific and Parallel Computing PARA 2010 : State of the Art in Scientific and Parallel Computing https://hal-supelec.archives-ouvertes.fr/hal-00491976 PARA 2010 : State of the Art in Scientific and Parallel Computing, Jun 2010, Reykjavik, Iceland. 4 p Parallelism GPGPU Asynchronism Scientific computing [INFO.INFO-DC]Computer Science [cs]/Distributed Parallel and Cluster Computing [cs.DC] info:eu-repo/semantics/conferenceObject Conference papers 2010 ftccsdartic 2021-12-12T05:59:14Z International audience We study the impact of asynchronism on parallel iterative algorithms in the particular context of local clusters of workstations including GPUs. The application test is a classical PDE problem of advection-diffusion-reaction in 3D. We propose an asynchronous version of a previously developed PDE solver using GPUs for the inner computations. The algorithm is tested with two kinds of clusters, a homogeneous one and a heterogeneous one. Conference Object Iceland Archive ouverte HAL (Hyper Article en Ligne, CCSD - Centre pour la Communication Scientifique Directe)
institution Open Polar
collection Archive ouverte HAL (Hyper Article en Ligne, CCSD - Centre pour la Communication Scientifique Directe)
op_collection_id ftccsdartic
language English
topic Parallelism
GPGPU
Asynchronism
Scientific computing
[INFO.INFO-DC]Computer Science [cs]/Distributed
Parallel
and Cluster Computing [cs.DC]
spellingShingle Parallelism
GPGPU
Asynchronism
Scientific computing
[INFO.INFO-DC]Computer Science [cs]/Distributed
Parallel
and Cluster Computing [cs.DC]
Contassot-Vivier, Sylvain
Jost, Thomas
Vialle, Stéphane
Impact of asynchronism on GPU accelerated parallel iterative computations
topic_facet Parallelism
GPGPU
Asynchronism
Scientific computing
[INFO.INFO-DC]Computer Science [cs]/Distributed
Parallel
and Cluster Computing [cs.DC]
description International audience We study the impact of asynchronism on parallel iterative algorithms in the particular context of local clusters of workstations including GPUs. The application test is a classical PDE problem of advection-diffusion-reaction in 3D. We propose an asynchronous version of a previously developed PDE solver using GPUs for the inner computations. The algorithm is tested with two kinds of clusters, a homogeneous one and a heterogeneous one.
author2 Algorithms for the Grid (ALGORILLE)
INRIA Lorraine
Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA)
Institut National de Recherche en Informatique et en Automatique (Inria)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS)
SUPELEC-Campus Metz
Ecole Supérieure d'Electricité - SUPELEC (FRANCE)
format Conference Object
author Contassot-Vivier, Sylvain
Jost, Thomas
Vialle, Stéphane
author_facet Contassot-Vivier, Sylvain
Jost, Thomas
Vialle, Stéphane
author_sort Contassot-Vivier, Sylvain
title Impact of asynchronism on GPU accelerated parallel iterative computations
title_short Impact of asynchronism on GPU accelerated parallel iterative computations
title_full Impact of asynchronism on GPU accelerated parallel iterative computations
title_fullStr Impact of asynchronism on GPU accelerated parallel iterative computations
title_full_unstemmed Impact of asynchronism on GPU accelerated parallel iterative computations
title_sort impact of asynchronism on gpu accelerated parallel iterative computations
publisher HAL CCSD
publishDate 2010
url https://hal-supelec.archives-ouvertes.fr/hal-00491976
https://hal-supelec.archives-ouvertes.fr/hal-00491976/document
https://hal-supelec.archives-ouvertes.fr/hal-00491976/file/Supelec620.pdf
op_coverage Reykjavik, Iceland
genre Iceland
genre_facet Iceland
op_source Proceedings of PARA 2010 conference: State of the Art in Scientific and Parallel Computing
PARA 2010 : State of the Art in Scientific and Parallel Computing
https://hal-supelec.archives-ouvertes.fr/hal-00491976
PARA 2010 : State of the Art in Scientific and Parallel Computing, Jun 2010, Reykjavik, Iceland. 4 p
op_relation hal-00491976
https://hal-supelec.archives-ouvertes.fr/hal-00491976
https://hal-supelec.archives-ouvertes.fr/hal-00491976/document
https://hal-supelec.archives-ouvertes.fr/hal-00491976/file/Supelec620.pdf
op_rights info:eu-repo/semantics/OpenAccess
_version_ 1766038219534630912