Performance study of flow computation in PC cluster

Chih Jen Lin, Kuang Chuan Ou, Keh-Chin Chang

Research output: Contribution to journalArticle

Abstract

In order to overcome the large demand of computer memory and CPU time required for the complex flow computations, a homemade nine-node PC cluster which is constructed by nine personal computers bought in the usual market serves as the computational tool for flow simulation. This parallel computational platform is operated under the Gentoo Linux operating system on which the messages passing interface (MPI) libraries are installed. The test problems consist of 2D and 3D lid-driven cavity flows and are parallelized by domain decomposition method. Two domain decomposition methods based on ID and 2D topologies are investigated and their performances are compared with each other. Comparison of the speedup extents under various network environments is also performed. It is concluded that the choice of ID or 2D topology depends on the complexity of the physical problems to be solved. It is also found that the promotion of the network environment such as using the gigabit Ethernet to replace the fast Ethernet leads to better speedup in the parallel computation.

Original languageEnglish
Pages (from-to)109-120
Number of pages12
JournalHangkong Taikong ji Minhang Xuekan/Journal of Aeronautics, Astronautics and Aviation
Volume38 B
Issue number2
Publication statusPublished - 2006 Oct 1

Fingerprint

Domain decomposition methods
Ethernet
topology
Topology
decomposition
memory (computers)
cavity flow
Computer operating systems
Flow simulation
personal computers
Message passing
promotion
messages
Personal computers
Interfaces (computer)
Program processors
platforms
Data storage equipment
cavity
market

All Science Journal Classification (ASJC) codes

  • Aerospace Engineering
  • Space and Planetary Science

Cite this

@article{8908445cd1d84a1f9edd3af295cb66c5,
title = "Performance study of flow computation in PC cluster",
abstract = "In order to overcome the large demand of computer memory and CPU time required for the complex flow computations, a homemade nine-node PC cluster which is constructed by nine personal computers bought in the usual market serves as the computational tool for flow simulation. This parallel computational platform is operated under the Gentoo Linux operating system on which the messages passing interface (MPI) libraries are installed. The test problems consist of 2D and 3D lid-driven cavity flows and are parallelized by domain decomposition method. Two domain decomposition methods based on ID and 2D topologies are investigated and their performances are compared with each other. Comparison of the speedup extents under various network environments is also performed. It is concluded that the choice of ID or 2D topology depends on the complexity of the physical problems to be solved. It is also found that the promotion of the network environment such as using the gigabit Ethernet to replace the fast Ethernet leads to better speedup in the parallel computation.",
author = "Lin, {Chih Jen} and Ou, {Kuang Chuan} and Keh-Chin Chang",
year = "2006",
month = "10",
day = "1",
language = "English",
volume = "38 B",
pages = "109--120",
journal = "Zhongguo Hangkong Taikong Xuehui Huikan/Transactions of the Aeronautical and Astronautical Society of the Republic of China",
issn = "1022-0666",
publisher = "The Aeronautical and Astronautical Society of the Republic of China",
number = "2",

}

Performance study of flow computation in PC cluster. / Lin, Chih Jen; Ou, Kuang Chuan; Chang, Keh-Chin.

In: Hangkong Taikong ji Minhang Xuekan/Journal of Aeronautics, Astronautics and Aviation, Vol. 38 B, No. 2, 01.10.2006, p. 109-120.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Performance study of flow computation in PC cluster

AU - Lin, Chih Jen

AU - Ou, Kuang Chuan

AU - Chang, Keh-Chin

PY - 2006/10/1

Y1 - 2006/10/1

N2 - In order to overcome the large demand of computer memory and CPU time required for the complex flow computations, a homemade nine-node PC cluster which is constructed by nine personal computers bought in the usual market serves as the computational tool for flow simulation. This parallel computational platform is operated under the Gentoo Linux operating system on which the messages passing interface (MPI) libraries are installed. The test problems consist of 2D and 3D lid-driven cavity flows and are parallelized by domain decomposition method. Two domain decomposition methods based on ID and 2D topologies are investigated and their performances are compared with each other. Comparison of the speedup extents under various network environments is also performed. It is concluded that the choice of ID or 2D topology depends on the complexity of the physical problems to be solved. It is also found that the promotion of the network environment such as using the gigabit Ethernet to replace the fast Ethernet leads to better speedup in the parallel computation.

AB - In order to overcome the large demand of computer memory and CPU time required for the complex flow computations, a homemade nine-node PC cluster which is constructed by nine personal computers bought in the usual market serves as the computational tool for flow simulation. This parallel computational platform is operated under the Gentoo Linux operating system on which the messages passing interface (MPI) libraries are installed. The test problems consist of 2D and 3D lid-driven cavity flows and are parallelized by domain decomposition method. Two domain decomposition methods based on ID and 2D topologies are investigated and their performances are compared with each other. Comparison of the speedup extents under various network environments is also performed. It is concluded that the choice of ID or 2D topology depends on the complexity of the physical problems to be solved. It is also found that the promotion of the network environment such as using the gigabit Ethernet to replace the fast Ethernet leads to better speedup in the parallel computation.

UR - http://www.scopus.com/inward/record.url?scp=33845521032&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845521032&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:33845521032

VL - 38 B

SP - 109

EP - 120

JO - Zhongguo Hangkong Taikong Xuehui Huikan/Transactions of the Aeronautical and Astronautical Society of the Republic of China

JF - Zhongguo Hangkong Taikong Xuehui Huikan/Transactions of the Aeronautical and Astronautical Society of the Republic of China

SN - 1022-0666

IS - 2

ER -