Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing

Takeshi Nanri, Takeshi Soga, Koji Kurihara, Feng Long Gu, Hiroaki Ishihata, Kazuaki Murakami

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

Abstract

This paper proposes and evaluates an algorithm of parallel matrix multiplication on compressed sparse-matrices. In addition to that, to overcome the problem of imbalanced distribution of non-zero elements over processors, a technology for dynamic load-balancing is examined. Measurements of the performance show that the algorithm achieves sufficient speed up. The experiments in this paper could not show the significant effect of the dynamic load-balancing. It requires more analysis to determine the reason of this.

Original languageEnglish
Title of host publicationComputation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007)
Pages106-109
Number of pages4
Edition2
DOIs
Publication statusPublished - Dec 1 2007
EventInternational Conference on Computational Methods in Science and Engineering 2007, ICCMSE 2007 - Corfu, Greece
Duration: Sep 25 2007Sep 30 2007

Publication series

NameAIP Conference Proceedings
Number2
Volume963
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Other

OtherInternational Conference on Computational Methods in Science and Engineering 2007, ICCMSE 2007
CountryGreece
CityCorfu
Period9/25/079/30/07

Fingerprint

dynamic loads
multiplication
evaluation
matrices
central processing units

All Science Journal Classification (ASJC) codes

  • Ecology, Evolution, Behavior and Systematics
  • Ecology
  • Plant Science
  • Physics and Astronomy(all)
  • Nature and Landscape Conservation

Cite this

Nanri, T., Soga, T., Kurihara, K., Gu, F. L., Ishihata, H., & Murakami, K. (2007). Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing. In Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007) (2 ed., pp. 106-109). (AIP Conference Proceedings; Vol. 963, No. 2). https://doi.org/10.1063/1.2835926

Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing. / Nanri, Takeshi; Soga, Takeshi; Kurihara, Koji; Gu, Feng Long; Ishihata, Hiroaki; Murakami, Kazuaki.

Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007). 2. ed. 2007. p. 106-109 (AIP Conference Proceedings; Vol. 963, No. 2).

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

Nanri, T, Soga, T, Kurihara, K, Gu, FL, Ishihata, H & Murakami, K 2007, Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing. in Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007). 2 edn, AIP Conference Proceedings, no. 2, vol. 963, pp. 106-109, International Conference on Computational Methods in Science and Engineering 2007, ICCMSE 2007, Corfu, Greece, 9/25/07. https://doi.org/10.1063/1.2835926
Nanri T, Soga T, Kurihara K, Gu FL, Ishihata H, Murakami K. Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing. In Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007). 2 ed. 2007. p. 106-109. (AIP Conference Proceedings; 2). https://doi.org/10.1063/1.2835926
Nanri, Takeshi ; Soga, Takeshi ; Kurihara, Koji ; Gu, Feng Long ; Ishihata, Hiroaki ; Murakami, Kazuaki. / Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing. Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007). 2. ed. 2007. pp. 106-109 (AIP Conference Proceedings; 2).
@inproceedings{0a6568c049c9490bb24e3f9515a6e327,
title = "Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing",
abstract = "This paper proposes and evaluates an algorithm of parallel matrix multiplication on compressed sparse-matrices. In addition to that, to overcome the problem of imbalanced distribution of non-zero elements over processors, a technology for dynamic load-balancing is examined. Measurements of the performance show that the algorithm achieves sufficient speed up. The experiments in this paper could not show the significant effect of the dynamic load-balancing. It requires more analysis to determine the reason of this.",
author = "Takeshi Nanri and Takeshi Soga and Koji Kurihara and Gu, {Feng Long} and Hiroaki Ishihata and Kazuaki Murakami",
year = "2007",
month = "12",
day = "1",
doi = "10.1063/1.2835926",
language = "English",
isbn = "9780735404786",
series = "AIP Conference Proceedings",
number = "2",
pages = "106--109",
booktitle = "Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007)",
edition = "2",

}

TY - GEN

T1 - Evaluation of the performance of parallel sparse-matrix multiplication and the effect of dynamic load-balancing

AU - Nanri, Takeshi

AU - Soga, Takeshi

AU - Kurihara, Koji

AU - Gu, Feng Long

AU - Ishihata, Hiroaki

AU - Murakami, Kazuaki

PY - 2007/12/1

Y1 - 2007/12/1

N2 - This paper proposes and evaluates an algorithm of parallel matrix multiplication on compressed sparse-matrices. In addition to that, to overcome the problem of imbalanced distribution of non-zero elements over processors, a technology for dynamic load-balancing is examined. Measurements of the performance show that the algorithm achieves sufficient speed up. The experiments in this paper could not show the significant effect of the dynamic load-balancing. It requires more analysis to determine the reason of this.

AB - This paper proposes and evaluates an algorithm of parallel matrix multiplication on compressed sparse-matrices. In addition to that, to overcome the problem of imbalanced distribution of non-zero elements over processors, a technology for dynamic load-balancing is examined. Measurements of the performance show that the algorithm achieves sufficient speed up. The experiments in this paper could not show the significant effect of the dynamic load-balancing. It requires more analysis to determine the reason of this.

UR - http://www.scopus.com/inward/record.url?scp=71449090722&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=71449090722&partnerID=8YFLogxK

U2 - 10.1063/1.2835926

DO - 10.1063/1.2835926

M3 - Conference contribution

AN - SCOPUS:71449090722

SN - 9780735404786

T3 - AIP Conference Proceedings

SP - 106

EP - 109

BT - Computation in Modern Science and Engineering - Proceedings of the International Conference on Computational Methods in Science and Engineering 2007 (ICCMSE 2007)

ER -