Show simple item record

dc.contributor.authorBueno Hedo, Javier
dc.contributor.authorPlanas, Judit
dc.contributor.authorDuran González, Alejandro
dc.contributor.authorBadia Sala, Rosa Maria
dc.contributor.authorMartorell Bofill, Xavier
dc.contributor.authorAyguadé Parra, Eduard
dc.contributor.authorLabarta Mancho, Jesús José
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
dc.date.accessioned2012-10-17T10:52:53Z
dc.date.available2012-10-17T10:57:37Z
dc.date.created2012
dc.date.issued2012
dc.identifier.citationBueno, J. [et al.]. Productive programming of GPU clusters with OmpSs. A: IEEE International Parallel and Distributed Processing Symposium. "2012 IEEE 26th International Parallel & Distributed Processing Symposium (IPDPS) 21-25 May 2012: Shanghai, China". Shanghai: 2012, p. 557-568.
dc.identifier.isbn978-1-4673-0975-2
dc.identifier.urihttp://hdl.handle.net/2117/16739
dc.description.abstractClusters of GPUs are emerging as a new computational scenario. Programming them requires the use of hybrid models that increase the complexity of the applications, reducing the productivity of programmers. We present the implementation of OmpSs for clusters of GPUs, which supports asynchrony and heterogeneity for task parallelism. It is based on annotating a serial application with directives that are translated by the compiler. With it, the same program that runs sequentially in a node with a single GPU can run in parallel in multiple GPUs either local (single node) or remote (cluster of GPUs). Besides performing a task-based parallelization, the runtime system moves the data as needed between the different nodes and GPUs minimizing the impact of communication by using affinity scheduling, caching, and by overlapping communication with the computational task. We show several applicactions programmed with OmpSs and their performance with multiple GPUs in a local node and in remote nodes. The results show good tradeoff between performance and effort from the programmer.
dc.format.extent12 p.
dc.language.isoeng
dc.subjectÀrees temàtiques de la UPC::Informàtica::Arquitectura de computadors::Arquitectures distribuïdes
dc.subject.lcshComputational grids (Computer systems)
dc.titleProductive programming of GPU clusters with OmpSs
dc.typeConference lecture
dc.subject.lemacComputació distribuïda
dc.contributor.groupUniversitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
dc.identifier.doi10.1109/IPDPS.2012.58
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttp://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6267858
dc.rights.accessRestricted access - publisher's policy
local.identifier.drac10962782
dc.description.versionPostprint (published version)
dc.relation.projectidinfo:eu-repo/grantAgreement/EC/FP7/248647/EU/ENabling technologies for a programmable many-CORE/ENCORE
dc.date.lift10000-01-01
local.citation.authorBueno, J.; Planas, J.; Duran, A.; Badia, R.; Martorell, X.; Ayguade, E.; Labarta, J.
local.citation.contributorIEEE International Parallel and Distributed Processing Symposium
local.citation.pubplaceShanghai
local.citation.publicationName2012 IEEE 26th International Parallel & Distributed Processing Symposium (IPDPS) 21-25 May 2012: Shanghai, China
local.citation.startingPage557
local.citation.endingPage568


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder