Transparently optimizating GNU/OCTAVE with GPGPU

This paper present an alternative to optimize GNU/Octave Numerical Methods, in a transparent way to the programer user. For this we used GPU computing with a NVIDIA’s graphics card through the CUDA toolkit, overlapped the BLAS library and also few OCT extensions.

Detalles Bibliográficos
Autores: Tibola, André Luis, Oliveira, Vinícius Menezes de
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2010
País:Brasil
Institución:Universidade Federal do Rio Grande (FURG)
Repositorio:Vetor (Online)
Idioma:portugués
OAI Identifier:oai:periodicos.furg.br:article/1744
Acceso en línea:https://periodicos.furg.br/vetor/article/view/1744
Access Level:acceso abierto
Palabra clave:BLAS
Computação de alto desempenho
CUDA
GPGPU
GNU/Octave
Nvidia
Descripción
Sumario:This paper present an alternative to optimize GNU/Octave Numerical Methods, in a transparent way to the programer user. For this we used GPU computing with a NVIDIA’s graphics card through the CUDA toolkit, overlapped the BLAS library and also few OCT extensions.