Afficher la notice abrégée

hal.structure.identifierEfficient runtime systems for parallel architectures [RUNTIME]
hal.structure.identifierLaboratoire Bordelais de Recherche en Informatique [LaBRI]
dc.contributor.authorTRAHAY, François
hal.structure.identifierEfficient runtime systems for parallel architectures [RUNTIME]
hal.structure.identifierLaboratoire Bordelais de Recherche en Informatique [LaBRI]
dc.contributor.authorDENIS, Alexandre
hal.structure.identifierEfficient runtime systems for parallel architectures [RUNTIME]
hal.structure.identifierLaboratoire Bordelais de Recherche en Informatique [LaBRI]
dc.contributor.authorAUMAGE, Olivier
hal.structure.identifierEfficient runtime systems for parallel architectures [RUNTIME]
hal.structure.identifierLaboratoire Bordelais de Recherche en Informatique [LaBRI]
dc.contributor.authorNAMYST, Raymond
dc.contributor.editorCapello
dc.contributor.editorHerault
dc.contributor.editorDongarra (Eds)
dc.date.accessioned2024-04-15T09:56:31Z
dc.date.available2024-04-15T09:56:31Z
dc.date.issued2007
dc.date.conference2007-10
dc.identifier.urihttps://oskar-bordeaux.fr/handle/20.500.12278/198841
dc.description.abstractEnMPI applications may waste thousands of CPU cycles if they do not efficiently overlap communications and computation. In this paper, we present a generic and portable I/O manager that is able to make communication progress asynchronously using tasklets. It chooses automatically the most appropriate communication method, depending on the context: multi-threaded application or not, SMP machine or not. We have implemented and evaluated our I/O manager with Mad-MPI, our own MPI implementation, and compared it to other existing MPI implementations regarding the ability to efficiently overlap communication and computation.
dc.language.isoen
dc.publisherSpringer
dc.subject.enPolling
dc.subject.enInterrupt
dc.subject.enThread
dc.subject.enScheduler
dc.subject.enHigh-Speed Network
dc.subject.enPioman
dc.subject.enNewMadeleine
dc.title.enImproving Reactivity and Communication Overlap in MPI using a Generic I/O Manager
dc.typeCommunication dans un congrès
dc.identifier.doi10.1007/978-3-540-75416-9_27
dc.subject.halInformatique [cs]/Calcul parallèle, distribué et partagé [cs.DC]
bordeaux.page170-177
bordeaux.volume4757
bordeaux.hal.laboratoriesLaboratoire Bordelais de Recherche en Informatique (LaBRI) - UMR 5800*
bordeaux.institutionUniversité de Bordeaux
bordeaux.institutionBordeaux INP
bordeaux.institutionCNRS
bordeaux.conference.titleEuroPVM/MPI 2007
bordeaux.countryFR
bordeaux.conference.cityParis
bordeaux.peerReviewedoui
hal.identifierinria-00177167
hal.version1
hal.invitednon
hal.proceedingsoui
hal.popularnon
hal.audienceInternationale
hal.origin.linkhttps://hal.archives-ouvertes.fr//inria-00177167v1
bordeaux.COinSctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.date=2007&rft.volume=4757&rft.spage=170-177&rft.epage=170-177&rft.au=TRAHAY,%20Fran%C3%A7ois&DENIS,%20Alexandre&AUMAGE,%20Olivier&NAMYST,%20Raymond&rft.genre=unknown


Fichier(s) constituant ce document

FichiersTailleFormatVue

Il n'y a pas de fichiers associés à ce document.

Ce document figure dans la(les) collection(s) suivante(s)

Afficher la notice abrégée