Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach

A wide range of applications in engineering and scientific computing are involved in the acceleration of the sparse matrix vector product (SpMV). Graphics Processing Units (GPUs) have recently emerged as platforms that yield outstanding acceleration factors. SpMV implementations for GPUs have alread...

Descripción completa

Detalles Bibliográficos
Autores: Vázquez, Francisco, Fernández, José Jesús, Garzón, Ester M.
Tipo de recurso: artículo
Estado:Versión enviada para evaluación y publicación
Fecha de publicación:2012
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/380549
Acceso en línea:http://hdl.handle.net/10261/380549
Access Level:acceso abierto
Palabra clave:Sparse matrix vector product
GPU computing
GPU performance modeling
id ES_3441dcd5d8ddb517e59b2c92a88be721
oai_identifier_str oai:digital.csic.es:10261/380549
network_acronym_str ES
network_name_str España
repository_id_str
spelling Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approachVázquez, FranciscoFernández, José JesúsGarzón, Ester M.Sparse matrix vector productGPU computingGPU performance modelingA wide range of applications in engineering and scientific computing are involved in the acceleration of the sparse matrix vector product (SpMV). Graphics Processing Units (GPUs) have recently emerged as platforms that yield outstanding acceleration factors. SpMV implementations for GPUs have already appeared on the scene. This work is focused on the ELLR-T algorithm to compute SpMV on GPU architecture, its performance is strongly dependent on the optimum selection of two parameters. Therefore, taking account that the memory operations dominate the performance of ELLR-T, an analytical model is proposed in order to obtain the auto-tuning of ELLR-T for particular combinations of sparse matrix and GPU architecture. The evaluation results with a representative set of test matrices show that the average performance achieved by auto-tuned ELLR-T by means of the proposed model is near to the optimum. A comparative analysis of ELLR-T against a variety of previous proposals shows that ELLR-T with the estimated configuration reaches the best performance on GPU architecture for the representative set of test matrices.This work has been funded by grants from the Spanish Ministry of Science and Innovation (TIN2008-01117), Junta de Andalucia (JA-P10-TIC-6002, JA-P08-TIC-3518) and Consejo Superior de Investigaciones Cientificas (CSIC-PIE-200920I075).Peer reviewedElsevierMinisterio de Ciencia e Innovación (España)Junta de AndalucíaConsejo Superior de Investigaciones Científicas (España)Consejo Superior de Investigaciones Científicas [https://ror.org/02gfc7t72]202520252012info:eu-repo/semantics/articlehttp://purl.org/coar/resource_type/c_6501Preprintinfo:eu-repo/semantics/submittedVersionapplication/pdfhttp://hdl.handle.net/10261/380549reponame:DIGITAL.CSIC. Repositorio Institucional del CSICinstname:Consejo Superior de Investigaciones Científicas (CSIC)Ingléshttps://doi.org/10.1016/j.parco.2011.08.003Síinfo:eu-repo/semantics/openAccessoai:digital.csic.es:10261/3805492026-05-22T06:33:51Z
dc.title.none.fl_str_mv Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
title Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
spellingShingle Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
Vázquez, Francisco
Sparse matrix vector product
GPU computing
GPU performance modeling
title_short Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
title_full Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
title_fullStr Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
title_full_unstemmed Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
title_sort Automatic tuning of the sparse matrix vector product on GPUs based on the ELLR-T approach
dc.creator.none.fl_str_mv Vázquez, Francisco
Fernández, José Jesús
Garzón, Ester M.
author Vázquez, Francisco
author_facet Vázquez, Francisco
Fernández, José Jesús
Garzón, Ester M.
author_role author
author2 Fernández, José Jesús
Garzón, Ester M.
author2_role author
author
dc.contributor.none.fl_str_mv Ministerio de Ciencia e Innovación (España)
Junta de Andalucía
Consejo Superior de Investigaciones Científicas (España)
Consejo Superior de Investigaciones Científicas [https://ror.org/02gfc7t72]
dc.subject.none.fl_str_mv Sparse matrix vector product
GPU computing
GPU performance modeling
topic Sparse matrix vector product
GPU computing
GPU performance modeling
description A wide range of applications in engineering and scientific computing are involved in the acceleration of the sparse matrix vector product (SpMV). Graphics Processing Units (GPUs) have recently emerged as platforms that yield outstanding acceleration factors. SpMV implementations for GPUs have already appeared on the scene. This work is focused on the ELLR-T algorithm to compute SpMV on GPU architecture, its performance is strongly dependent on the optimum selection of two parameters. Therefore, taking account that the memory operations dominate the performance of ELLR-T, an analytical model is proposed in order to obtain the auto-tuning of ELLR-T for particular combinations of sparse matrix and GPU architecture. The evaluation results with a representative set of test matrices show that the average performance achieved by auto-tuned ELLR-T by means of the proposed model is near to the optimum. A comparative analysis of ELLR-T against a variety of previous proposals shows that ELLR-T with the estimated configuration reaches the best performance on GPU architecture for the representative set of test matrices.
publishDate 2012
dc.date.none.fl_str_mv 2012
2025
2025
dc.type.none.fl_str_mv info:eu-repo/semantics/article
http://purl.org/coar/resource_type/c_6501
Preprint
info:eu-repo/semantics/submittedVersion
format article
status_str submittedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/10261/380549
url http://hdl.handle.net/10261/380549
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv https://doi.org/10.1016/j.parco.2011.08.003
Sí
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:DIGITAL.CSIC. Repositorio Institucional del CSIC
instname:Consejo Superior de Investigaciones Científicas (CSIC)
instname_str Consejo Superior de Investigaciones Científicas (CSIC)
reponame_str DIGITAL.CSIC. Repositorio Institucional del CSIC
collection DIGITAL.CSIC. Repositorio Institucional del CSIC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869405799038582784
score 15.81155

Similares en LA Referencia