New approach for task clustering

Weiping Zhu, Tyng Yeu Liang, Ce Kuen Shieh

研究成果: Paper

摘要

Static clustering has been used to group tasks for parallel processing. Most clustering methods used for current multithreaded DSM systems only consider the workload balance. In contrast, in this paper we present a static method to cluster closely related tasks of an application onto a multithreaded DSM system. This method relies on the Hopfield neural network to find optimal or near-optimal clusters. An optimal solution identified by this method tends to minimize load imbalance and communication overhead. We have implemented this method on Cohesion which is a multithreaded DSM system. Three programs, SOR, Nbody, and Gaussian Elimination, are used to test the effectiveness of this method. The result shows that our method indeed can find optimal or near-optimal clustering for these programs.

原文English
頁面538-542
頁數5
出版狀態Published - 1998 一月 1
事件Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2) - Beijing, China
持續時間: 1997 十月 281997 十月 31

Other

OtherProceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2)
城市Beijing, China
期間97-10-2897-10-31

指紋

Hopfield neural networks
Communication
Processing

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Engineering(all)

引用此文

Zhu, W., Liang, T. Y., & Shieh, C. K. (1998). New approach for task clustering. 538-542. 論文發表於 Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2), Beijing, China, .
Zhu, Weiping ; Liang, Tyng Yeu ; Shieh, Ce Kuen. / New approach for task clustering. 論文發表於 Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2), Beijing, China, .5 p.
@conference{17d92b8a346344de9aabc01597593ee9,
title = "New approach for task clustering",
abstract = "Static clustering has been used to group tasks for parallel processing. Most clustering methods used for current multithreaded DSM systems only consider the workload balance. In contrast, in this paper we present a static method to cluster closely related tasks of an application onto a multithreaded DSM system. This method relies on the Hopfield neural network to find optimal or near-optimal clusters. An optimal solution identified by this method tends to minimize load imbalance and communication overhead. We have implemented this method on Cohesion which is a multithreaded DSM system. Three programs, SOR, Nbody, and Gaussian Elimination, are used to test the effectiveness of this method. The result shows that our method indeed can find optimal or near-optimal clustering for these programs.",
author = "Weiping Zhu and Liang, {Tyng Yeu} and Shieh, {Ce Kuen}",
year = "1998",
month = "1",
day = "1",
language = "English",
pages = "538--542",
note = "Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2) ; Conference date: 28-10-1997 Through 31-10-1997",

}

Zhu, W, Liang, TY & Shieh, CK 1998, 'New approach for task clustering', 論文發表於 Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2), Beijing, China, 97-10-28 - 97-10-31 頁 538-542.

New approach for task clustering. / Zhu, Weiping; Liang, Tyng Yeu; Shieh, Ce Kuen.

1998. 538-542 論文發表於 Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2), Beijing, China, .

研究成果: Paper

TY - CONF

T1 - New approach for task clustering

AU - Zhu, Weiping

AU - Liang, Tyng Yeu

AU - Shieh, Ce Kuen

PY - 1998/1/1

Y1 - 1998/1/1

N2 - Static clustering has been used to group tasks for parallel processing. Most clustering methods used for current multithreaded DSM systems only consider the workload balance. In contrast, in this paper we present a static method to cluster closely related tasks of an application onto a multithreaded DSM system. This method relies on the Hopfield neural network to find optimal or near-optimal clusters. An optimal solution identified by this method tends to minimize load imbalance and communication overhead. We have implemented this method on Cohesion which is a multithreaded DSM system. Three programs, SOR, Nbody, and Gaussian Elimination, are used to test the effectiveness of this method. The result shows that our method indeed can find optimal or near-optimal clustering for these programs.

AB - Static clustering has been used to group tasks for parallel processing. Most clustering methods used for current multithreaded DSM systems only consider the workload balance. In contrast, in this paper we present a static method to cluster closely related tasks of an application onto a multithreaded DSM system. This method relies on the Hopfield neural network to find optimal or near-optimal clusters. An optimal solution identified by this method tends to minimize load imbalance and communication overhead. We have implemented this method on Cohesion which is a multithreaded DSM system. Three programs, SOR, Nbody, and Gaussian Elimination, are used to test the effectiveness of this method. The result shows that our method indeed can find optimal or near-optimal clustering for these programs.

UR - http://www.scopus.com/inward/record.url?scp=0031681628&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031681628&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0031681628

SP - 538

EP - 542

ER -

Zhu W, Liang TY, Shieh CK. New approach for task clustering. 1998. 論文發表於 Proceedings of the 1997 IEEE International Conference on Intelligent Processing Systems, ICIPS'97. Part 1 (of 2), Beijing, China, .