SSG-AT: An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam

Satoshi Ito, Satoshi Ohshima, Takahiro Katagiri

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

We are developing ppOpen-AT, which is an infrastructureof auto-tuning (AT) for ppOpen-HPC. ppOpen-HPC is numerical middleware for post Petascale era. In this study, we propose a new auto-tuning (AT) facility for semi-structured grids in OpenFOAM. We focus on sparse matrix-vector multiplication and the matrix storage formats. Using the features of input data and mesh connectivity, we propose a hybrid storage format that is suitable for semistructured grids. We evaluate the proposed AT facility on the T2K supercomputer and an Intel Xeon cluster. For a typical computational fluid dynamics scenario, we obtain speedup factors of 1.3 on the T2K and 1.84 on the Xeon cluster. These results indicate that the proposed AT method has the potential to select the optimal data format according to features of the input sparse matrix.

Original languageEnglish
Pages191-197
Number of pages7
DOIs
Publication statusPublished - Dec 1 2012
Event2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012 - Aizu-Wakamatsu, Fukushima, Japan
Duration: Sep 20 2012Sep 22 2012

Other

Other2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012
CountryJapan
CityAizu-Wakamatsu, Fukushima
Period9/20/129/22/12

Fingerprint

Tuning
Supercomputers
Middleware
Computational fluid dynamics

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Cite this

Ito, S., Ohshima, S., & Katagiri, T. (2012). SSG-AT: An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam. 191-197. Paper presented at 2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012, Aizu-Wakamatsu, Fukushima, Japan. https://doi.org/10.1109/MCSoC.2012.26

SSG-AT : An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam. / Ito, Satoshi; Ohshima, Satoshi; Katagiri, Takahiro.

2012. 191-197 Paper presented at 2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012, Aizu-Wakamatsu, Fukushima, Japan.

Research output: Contribution to conferencePaper

Ito, S, Ohshima, S & Katagiri, T 2012, 'SSG-AT: An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam' Paper presented at 2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012, Aizu-Wakamatsu, Fukushima, Japan, 9/20/12 - 9/22/12, pp. 191-197. https://doi.org/10.1109/MCSoC.2012.26
Ito S, Ohshima S, Katagiri T. SSG-AT: An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam. 2012. Paper presented at 2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012, Aizu-Wakamatsu, Fukushima, Japan. https://doi.org/10.1109/MCSoC.2012.26
Ito, Satoshi ; Ohshima, Satoshi ; Katagiri, Takahiro. / SSG-AT : An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam. Paper presented at 2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012, Aizu-Wakamatsu, Fukushima, Japan.7 p.
@conference{92a40c6376f940cd870a86eb0de85e3a,
title = "SSG-AT: An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam",
abstract = "We are developing ppOpen-AT, which is an infrastructureof auto-tuning (AT) for ppOpen-HPC. ppOpen-HPC is numerical middleware for post Petascale era. In this study, we propose a new auto-tuning (AT) facility for semi-structured grids in OpenFOAM. We focus on sparse matrix-vector multiplication and the matrix storage formats. Using the features of input data and mesh connectivity, we propose a hybrid storage format that is suitable for semistructured grids. We evaluate the proposed AT facility on the T2K supercomputer and an Intel Xeon cluster. For a typical computational fluid dynamics scenario, we obtain speedup factors of 1.3 on the T2K and 1.84 on the Xeon cluster. These results indicate that the proposed AT method has the potential to select the optimal data format according to features of the input sparse matrix.",
author = "Satoshi Ito and Satoshi Ohshima and Takahiro Katagiri",
year = "2012",
month = "12",
day = "1",
doi = "10.1109/MCSoC.2012.26",
language = "English",
pages = "191--197",
note = "2012 IEEE 6th International Symposium on Embedded Multi-Core Systems on Chips, MCSoC 2012 ; Conference date: 20-09-2012 Through 22-09-2012",

}

TY - CONF

T1 - SSG-AT

T2 - An auto-tuning method of sparse matrix-vector multiplicataion for semi-structured grids-An adaptation to openfoam

AU - Ito, Satoshi

AU - Ohshima, Satoshi

AU - Katagiri, Takahiro

PY - 2012/12/1

Y1 - 2012/12/1

N2 - We are developing ppOpen-AT, which is an infrastructureof auto-tuning (AT) for ppOpen-HPC. ppOpen-HPC is numerical middleware for post Petascale era. In this study, we propose a new auto-tuning (AT) facility for semi-structured grids in OpenFOAM. We focus on sparse matrix-vector multiplication and the matrix storage formats. Using the features of input data and mesh connectivity, we propose a hybrid storage format that is suitable for semistructured grids. We evaluate the proposed AT facility on the T2K supercomputer and an Intel Xeon cluster. For a typical computational fluid dynamics scenario, we obtain speedup factors of 1.3 on the T2K and 1.84 on the Xeon cluster. These results indicate that the proposed AT method has the potential to select the optimal data format according to features of the input sparse matrix.

AB - We are developing ppOpen-AT, which is an infrastructureof auto-tuning (AT) for ppOpen-HPC. ppOpen-HPC is numerical middleware for post Petascale era. In this study, we propose a new auto-tuning (AT) facility for semi-structured grids in OpenFOAM. We focus on sparse matrix-vector multiplication and the matrix storage formats. Using the features of input data and mesh connectivity, we propose a hybrid storage format that is suitable for semistructured grids. We evaluate the proposed AT facility on the T2K supercomputer and an Intel Xeon cluster. For a typical computational fluid dynamics scenario, we obtain speedup factors of 1.3 on the T2K and 1.84 on the Xeon cluster. These results indicate that the proposed AT method has the potential to select the optimal data format according to features of the input sparse matrix.

UR - http://www.scopus.com/inward/record.url?scp=84872574612&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872574612&partnerID=8YFLogxK

U2 - 10.1109/MCSoC.2012.26

DO - 10.1109/MCSoC.2012.26

M3 - Paper

AN - SCOPUS:84872574612

SP - 191

EP - 197

ER -