<link rel="stylesheet" href="styles.f3b1fba60ec7970c.css">

Análisis de paralelización con memoria compartida y memoria distribuida en clusters de nodos con múltiples núcleos

cic.isFulltexttruees
cic.isPeerReviewedtruees
cic.lugarDesarrolloInstituto de Investigación en Informática es
cic.parentTypeObjeto de conferencia
cic.versioninfo:eu-repo/semantics/submittedVersiones
dc.date.accessioned2016-08-10T12:29:57Z
dc.date.available2016-08-10T12:29:57Z
dc.identifier.urihttps://digital.cic.gba.gob.ar/handle/11746/3504
dc.titleAnálisis de paralelización con memoria compartida y memoria distribuida en clusters de nodos con múltiples núcleoses
dc.typeDocumento de conferenciaes
dcterms.abstractThis article presents the alternatives and performance results obtained after analyzing parallelization alternatives in clusters of nodes with multiple cores. The ultimate goal is to show if both processing and parallelization models (shared memory and distributed memory) need to be taken into account, or if only one of them is enough. The application used is classical in the context of highperformance computing: matrix multiplication. Even though this operation is representative of linear algebra applications, results are shown in terms of the conditions under which performance can be optimized and where algorithm parallelization efforts should be focused on for clusters of nodes with multiple cores. These clusters are nowadays considered as low-cost standards, since almost any desktop computer used to build clusters is based on a multi-core processor, and even on multi-processors. In any case, all processing units should be used to their maximum to optimize the performance of parallel applications.en
dcterms.creator.authorWolfmann, Aaron Gustavo Horacioes
dcterms.creator.authorTinetti, Fernando Gustavoes
dcterms.descriptionWorkshop de Procesamiento Distribuido y Paralelo (WPDP)es
dcterms.extent12 p.es
dcterms.identifier.urlRecurso Completoes
dcterms.isPartOf.issueXIV Congreso Argentino de Ciencias de la Computaciónes
dcterms.isPartOf.seriesCongreso Argentino de Ciencias de la Computaciónes
dcterms.issued2008-10
dcterms.languageEspañoles
dcterms.licenseAttribution 4.0 International (BY 4.0)es
dcterms.subjectShared memoryen
dcterms.subjectDistributed memoriesen
dcterms.subjectParallelen
dcterms.subject.materiaCiencias Informáticases

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Tinetti - Análisis de paralelización con memoria compartida.pdf-PDFA.pdf
Size:
190.6 KB
Format:
Adobe Portable Document Format
Description:
Documento completo