Improving the performance of matrix inversion with a Tesla GPU

We study two different techniques for the computation of a matrix inverse, the traditional approach based on Gaussian factorization and the Gauss-Jordan elimination alternative more suitable for parallel architectures. The target architecture is a current general-purpose multi-core processor (CPU) c...

Descripción completa

Detalles Bibliográficos
Autores principales: Ezzatti, Pablo, Quintana Ortí, Enrique S., Remón, Alfredo
Formato: Objeto de conferencia
Lenguaje:Inglés
Publicado: 2010
Materias:
Ciencias Informáticas
GPU
CPU
Efficiency
Acceso en línea:http://sedici.unlp.edu.ar/handle/10915/152637
http://39jaiio.sadio.org.ar/sites/default/files/39jaiio-hpc-03.pdf
Aporte de:
SEDICI (UNLP) de Universidad Nacional de La Plata
id I19-R120-10915-152637
record_format dspace
spelling I19-R120-10915-1526372023-05-09T20:04:13Z http://sedici.unlp.edu.ar/handle/10915/152637 http://39jaiio.sadio.org.ar/sites/default/files/39jaiio-hpc-03.pdf issn:1851-9326 Improving the performance of matrix inversion with a Tesla GPU Ezzatti, Pablo Quintana Ortí, Enrique S. Remón, Alfredo 2010 2010 2023-05-09T13:55:08Z en Ciencias Informáticas GPU CPU Efficiency We study two different techniques for the computation of a matrix inverse, the traditional approach based on Gaussian factorization and the Gauss-Jordan elimination alternative more suitable for parallel architectures. The target architecture is a current general-purpose multi-core processor (CPU) connected to a graphics processor (GPU). Parallelism is obtained from the use of libraries MKL (for the CPU) and CUBLAS (for the GPU), as well as, performing simultaneously operations in both architectures. Numerical experiments performed on a system equipped with two Intel QuadCore processors and a Tesla C1060 GPU, illustrate the efficiency attained by the Gauss-Jordan elimination implementation. Sociedad Argentina de Informática e Investigación Operativa Objeto de conferencia Objeto de conferencia http://creativecommons.org/licenses/by-nc-sa/4.0/ Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) application/pdf 3211-3219
institution Universidad Nacional de La Plata
institution_str I-19
repository_str R-120
collection SEDICI (UNLP)
language Inglés
topic Ciencias Informáticas
GPU
CPU
Efficiency
spellingShingle Ciencias Informáticas
GPU
CPU
Efficiency
Ezzatti, Pablo
Quintana Ortí, Enrique S.
Remón, Alfredo
Improving the performance of matrix inversion with a Tesla GPU
topic_facet Ciencias Informáticas
GPU
CPU
Efficiency
description We study two different techniques for the computation of a matrix inverse, the traditional approach based on Gaussian factorization and the Gauss-Jordan elimination alternative more suitable for parallel architectures. The target architecture is a current general-purpose multi-core processor (CPU) connected to a graphics processor (GPU). Parallelism is obtained from the use of libraries MKL (for the CPU) and CUBLAS (for the GPU), as well as, performing simultaneously operations in both architectures. Numerical experiments performed on a system equipped with two Intel QuadCore processors and a Tesla C1060 GPU, illustrate the efficiency attained by the Gauss-Jordan elimination implementation.
format Objeto de conferencia
Objeto de conferencia
author Ezzatti, Pablo
Quintana Ortí, Enrique S.
Remón, Alfredo
author_facet Ezzatti, Pablo
Quintana Ortí, Enrique S.
Remón, Alfredo
author_sort Ezzatti, Pablo
title Improving the performance of matrix inversion with a Tesla GPU
title_short Improving the performance of matrix inversion with a Tesla GPU
title_full Improving the performance of matrix inversion with a Tesla GPU
title_fullStr Improving the performance of matrix inversion with a Tesla GPU
title_full_unstemmed Improving the performance of matrix inversion with a Tesla GPU
title_sort improving the performance of matrix inversion with a tesla gpu
publishDate 2010
url http://sedici.unlp.edu.ar/handle/10915/152637
http://39jaiio.sadio.org.ar/sites/default/files/39jaiio-hpc-03.pdf
work_keys_str_mv AT ezzattipablo improvingtheperformanceofmatrixinversionwithateslagpu
AT quintanaortienriques improvingtheperformanceofmatrixinversionwithateslagpu
AT remonalfredo improvingtheperformanceofmatrixinversionwithateslagpu
_version_ 1765660136484896768

Ejemplares similares