Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs

Autores
Costanzo, Manuel; Rucci, Enzo; García-Sánchez, Carlos; Naiouf, Marcelo
Año de publicación
2023
Idioma
inglés
Tipo de recurso
documento de conferencia
Estado
versión publicada
Descripción
The heterogeneous computing paradigm has led to the need for portable and efficient programming solutions that can leverage the capabilities of various hardware devices, such as NVIDIA, Intel, and AMD GPUs. This study evaluates the performance and portability of the SYCL and CUDA languages for a matrix multiplication (MM) application across different GPU architectures. The experimental work showed that, while the CUDA implementation outperforms the SYCL implementation on NVIDIA devices due to optimizations provided by the nvcc compiler, the latter implementation demonstrated remarkable code portability to other GPU architectures, such as AMD and Intel. Furthermore, the architectural efficiency percentages obtained on AMD and Intel GPUs showed consistency with the results observed on NVIDIA devices.
Facultad de Informática
Materia
Ciencias Informáticas
oneAPI
SYCL
GPU
CUDA
Performance portability
Nivel de accesibilidad
acceso abierto
Condiciones de uso
http://creativecommons.org/licenses/by-nc-sa/4.0/
Repositorio
SEDICI (UNLP)
Institución
Universidad Nacional de La Plata
OAI Identificador
oai:sedici.unlp.edu.ar:10915/155420

id SEDICI_bf452f6a5ff30203bb3f01d89ac48f73
oai_identifier_str oai:sedici.unlp.edu.ar:10915/155420
network_acronym_str SEDICI
repository_id_str 1329
network_name_str SEDICI (UNLP)
spelling Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUsCostanzo, ManuelRucci, EnzoGarcía-Sánchez, CarlosNaiouf, MarceloCiencias InformáticasoneAPISYCLGPUCUDAPerformance portabilityThe heterogeneous computing paradigm has led to the need for portable and efficient programming solutions that can leverage the capabilities of various hardware devices, such as NVIDIA, Intel, and AMD GPUs. This study evaluates the performance and portability of the SYCL and CUDA languages for a matrix multiplication (MM) application across different GPU architectures. The experimental work showed that, while the CUDA implementation outperforms the SYCL implementation on NVIDIA devices due to optimizations provided by the nvcc compiler, the latter implementation demonstrated remarkable code portability to other GPU architectures, such as AMD and Intel. Furthermore, the architectural efficiency percentages obtained on AMD and Intel GPUs showed consistency with the results observed on NVIDIA devices.Facultad de Informática2023-06info:eu-repo/semantics/conferenceObjectinfo:eu-repo/semantics/publishedVersionObjeto de conferenciahttp://purl.org/coar/resource_type/c_5794info:ar-repo/semantics/documentoDeConferenciaapplication/pdf13-18http://sedici.unlp.edu.ar/handle/10915/155420enginfo:eu-repo/semantics/altIdentifier/isbn/978-950-34-2271-7info:eu-repo/semantics/reference/hdl/10915/155281info:eu-repo/semantics/openAccesshttp://creativecommons.org/licenses/by-nc-sa/4.0/Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)reponame:SEDICI (UNLP)instname:Universidad Nacional de La Platainstacron:UNLP2025-09-29T11:40:21Zoai:sedici.unlp.edu.ar:10915/155420Institucionalhttp://sedici.unlp.edu.ar/Universidad públicaNo correspondehttp://sedici.unlp.edu.ar/oai/snrdalira@sedici.unlp.edu.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:13292025-09-29 11:40:21.525SEDICI (UNLP) - Universidad Nacional de La Platafalse
dc.title.none.fl_str_mv Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
title Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
spellingShingle Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
Costanzo, Manuel
Ciencias Informáticas
oneAPI
SYCL
GPU
CUDA
Performance portability
title_short Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
title_full Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
title_fullStr Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
title_full_unstemmed Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
title_sort Brief performance portability analysis of a matrix multiplication kernel on multiple vendor GPUs
dc.creator.none.fl_str_mv Costanzo, Manuel
Rucci, Enzo
García-Sánchez, Carlos
Naiouf, Marcelo
author Costanzo, Manuel
author_facet Costanzo, Manuel
Rucci, Enzo
García-Sánchez, Carlos
Naiouf, Marcelo
author_role author
author2 Rucci, Enzo
García-Sánchez, Carlos
Naiouf, Marcelo
author2_role author
author
author
dc.subject.none.fl_str_mv Ciencias Informáticas
oneAPI
SYCL
GPU
CUDA
Performance portability
topic Ciencias Informáticas
oneAPI
SYCL
GPU
CUDA
Performance portability
dc.description.none.fl_txt_mv The heterogeneous computing paradigm has led to the need for portable and efficient programming solutions that can leverage the capabilities of various hardware devices, such as NVIDIA, Intel, and AMD GPUs. This study evaluates the performance and portability of the SYCL and CUDA languages for a matrix multiplication (MM) application across different GPU architectures. The experimental work showed that, while the CUDA implementation outperforms the SYCL implementation on NVIDIA devices due to optimizations provided by the nvcc compiler, the latter implementation demonstrated remarkable code portability to other GPU architectures, such as AMD and Intel. Furthermore, the architectural efficiency percentages obtained on AMD and Intel GPUs showed consistency with the results observed on NVIDIA devices.
Facultad de Informática
description The heterogeneous computing paradigm has led to the need for portable and efficient programming solutions that can leverage the capabilities of various hardware devices, such as NVIDIA, Intel, and AMD GPUs. This study evaluates the performance and portability of the SYCL and CUDA languages for a matrix multiplication (MM) application across different GPU architectures. The experimental work showed that, while the CUDA implementation outperforms the SYCL implementation on NVIDIA devices due to optimizations provided by the nvcc compiler, the latter implementation demonstrated remarkable code portability to other GPU architectures, such as AMD and Intel. Furthermore, the architectural efficiency percentages obtained on AMD and Intel GPUs showed consistency with the results observed on NVIDIA devices.
publishDate 2023
dc.date.none.fl_str_mv 2023-06
dc.type.none.fl_str_mv info:eu-repo/semantics/conferenceObject
info:eu-repo/semantics/publishedVersion
Objeto de conferencia
http://purl.org/coar/resource_type/c_5794
info:ar-repo/semantics/documentoDeConferencia
format conferenceObject
status_str publishedVersion
dc.identifier.none.fl_str_mv http://sedici.unlp.edu.ar/handle/10915/155420
url http://sedici.unlp.edu.ar/handle/10915/155420
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/isbn/978-950-34-2271-7
info:eu-repo/semantics/reference/hdl/10915/155281
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
eu_rights_str_mv openAccess
rights_invalid_str_mv http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
dc.format.none.fl_str_mv application/pdf
13-18
dc.source.none.fl_str_mv reponame:SEDICI (UNLP)
instname:Universidad Nacional de La Plata
instacron:UNLP
reponame_str SEDICI (UNLP)
collection SEDICI (UNLP)
instname_str Universidad Nacional de La Plata
instacron_str UNLP
institution UNLP
repository.name.fl_str_mv SEDICI (UNLP) - Universidad Nacional de La Plata
repository.mail.fl_str_mv alira@sedici.unlp.edu.ar
_version_ 1844616276531478528
score 13.070432