Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment

Satoshi Ohshima, Kenji Kise, Takahiro Katagiri, Toshitsugu Yuba

Research output: Chapter in Book/Report/Conference proceedingConference contribution

25 Citations (Scopus)

Abstract

GPUs for numerical computations are becoming an attractive alternative in research. In this paper, we propose a new parallel processing environment for matrix multiplications by using both CPUs and GPUs. The execution time of matrix multiplications can be decreased to 40.1% by our method, compared with using the fastest of either CPU only case or GPU only case. Our method performs well when matrix sizes are large.

Original languageEnglish
Title of host publicationHigh Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers
Pages305-318
Number of pages14
Publication statusPublished - Dec 1 2007
Event7th International Meeting on High-Performance Computing for Computational Science, VECPAR 2006 - Rio de Janeiro, Brazil
Duration: Jun 10 2006Jun 13 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4395 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other7th International Meeting on High-Performance Computing for Computational Science, VECPAR 2006
CountryBrazil
CityRio de Janeiro
Period6/10/066/13/06

Fingerprint

Heterogeneous Environment
Matrix multiplication
Parallel Processing
Program processors
Processing
Numerical Computation
Execution Time
Alternatives
Graphics processing unit

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Ohshima, S., Kise, K., Katagiri, T., & Yuba, T. (2007). Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment. In High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers (pp. 305-318). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4395 LNCS).

Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment. / Ohshima, Satoshi; Kise, Kenji; Katagiri, Takahiro; Yuba, Toshitsugu.

High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers. 2007. p. 305-318 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4395 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ohshima, S, Kise, K, Katagiri, T & Yuba, T 2007, Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment. in High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4395 LNCS, pp. 305-318, 7th International Meeting on High-Performance Computing for Computational Science, VECPAR 2006, Rio de Janeiro, Brazil, 6/10/06.
Ohshima S, Kise K, Katagiri T, Yuba T. Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment. In High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers. 2007. p. 305-318. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Ohshima, Satoshi ; Kise, Kenji ; Katagiri, Takahiro ; Yuba, Toshitsugu. / Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment. High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers. 2007. pp. 305-318 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{9be565d0ed6c4cc7b6f5f9fa828297e0,
title = "Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment",
abstract = "GPUs for numerical computations are becoming an attractive alternative in research. In this paper, we propose a new parallel processing environment for matrix multiplications by using both CPUs and GPUs. The execution time of matrix multiplications can be decreased to 40.1{\%} by our method, compared with using the fastest of either CPU only case or GPU only case. Our method performs well when matrix sizes are large.",
author = "Satoshi Ohshima and Kenji Kise and Takahiro Katagiri and Toshitsugu Yuba",
year = "2007",
month = "12",
day = "1",
language = "English",
isbn = "9783540713500",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "305--318",
booktitle = "High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers",

}

TY - GEN

T1 - Parallel processing of matrix multiplication in a CPU and GPU heterogeneous environment

AU - Ohshima, Satoshi

AU - Kise, Kenji

AU - Katagiri, Takahiro

AU - Yuba, Toshitsugu

PY - 2007/12/1

Y1 - 2007/12/1

N2 - GPUs for numerical computations are becoming an attractive alternative in research. In this paper, we propose a new parallel processing environment for matrix multiplications by using both CPUs and GPUs. The execution time of matrix multiplications can be decreased to 40.1% by our method, compared with using the fastest of either CPU only case or GPU only case. Our method performs well when matrix sizes are large.

AB - GPUs for numerical computations are becoming an attractive alternative in research. In this paper, we propose a new parallel processing environment for matrix multiplications by using both CPUs and GPUs. The execution time of matrix multiplications can be decreased to 40.1% by our method, compared with using the fastest of either CPU only case or GPU only case. Our method performs well when matrix sizes are large.

UR - http://www.scopus.com/inward/record.url?scp=38049125516&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38049125516&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:38049125516

SN - 9783540713500

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 305

EP - 318

BT - High Performance Computing for Computational Science - VECPAR 2006 - 7th International Conference, Revised Selected and Invited Papers

ER -