Preliminary investigation of distributed shared memory system on a cluster of high performance clusters

Takeshi Nanri, Yoshitaka Watanabe, Hiyoyuki Sato, Masaaki Shimasaki

研究成果: 著書/レポートタイプへの貢献会議での発言

抄録

This paper introduces design and basic performance of the DSM(distributed shared memory) system on a cluster of clusters. Networking devices such as Myrinet have improved the performance of cluster systems significantly. In addition to that, such kind of network devices introduced a new hierarchical architecture; a multi-cluster, a cluster of high-performance clusters. To ease the difficulty of programming with message passing, which is the conventional programming paradigm on cluster systems, many DSM (distributed shared memory) systems have been developed in recent years. However, there have been no DSM systems developed on multi-clusters. The DSM system consists of a runtime system to support basic functions for accessing virtual shared memory built on such environment. The functions are allocation of global data, read and write accesses to global data, synchronization of the whole system, and mutual exclusion. The authors have evaluated the performance of the runtime system, built on a SMP cluster, COMPaS, at RWCP(Real World Computing Partnership) in Tsukuba, Japan. The result shows that a read access to remote memory on the same cluster costs about 0.2msec, while a read access to remote memory on other cluster costs about 1.3msec. The execution time of LU decomposition on a multi-cluster consisting two clusters of three PCs is about 2.8times faster than the time on one PC.

元の言語英語
ホスト出版物のタイトルEuropean Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000
出版物ステータス出版済み - 12 1 2000
イベントEuropean Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000 - Barcelona, スペイン
継続期間: 9 11 20009 14 2000

その他

その他European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000
スペイン
Barcelona
期間9/11/009/14/00

Fingerprint

Distributed Shared Memory
High Performance
Data storage equipment
Computer systems
Runtime Systems
Message passing
Costs
Programming
Synchronization
Virtual Memory
LU decomposition
Decomposition
Mutual Exclusion
Message Passing
Shared Memory
Japan
Networking
Execution Time

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Applied Mathematics

これを引用

Nanri, T., Watanabe, Y., Sato, H., & Shimasaki, M. (2000). Preliminary investigation of distributed shared memory system on a cluster of high performance clusters. : European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000

Preliminary investigation of distributed shared memory system on a cluster of high performance clusters. / Nanri, Takeshi; Watanabe, Yoshitaka; Sato, Hiyoyuki; Shimasaki, Masaaki.

European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000. 2000.

研究成果: 著書/レポートタイプへの貢献会議での発言

Nanri, T, Watanabe, Y, Sato, H & Shimasaki, M 2000, Preliminary investigation of distributed shared memory system on a cluster of high performance clusters. : European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000. European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000, Barcelona, スペイン, 9/11/00.
Nanri T, Watanabe Y, Sato H, Shimasaki M. Preliminary investigation of distributed shared memory system on a cluster of high performance clusters. : European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000. 2000
Nanri, Takeshi ; Watanabe, Yoshitaka ; Sato, Hiyoyuki ; Shimasaki, Masaaki. / Preliminary investigation of distributed shared memory system on a cluster of high performance clusters. European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000. 2000.
@inproceedings{092a2897d45540dabedb98a26255eed6,
title = "Preliminary investigation of distributed shared memory system on a cluster of high performance clusters",
abstract = "This paper introduces design and basic performance of the DSM(distributed shared memory) system on a cluster of clusters. Networking devices such as Myrinet have improved the performance of cluster systems significantly. In addition to that, such kind of network devices introduced a new hierarchical architecture; a multi-cluster, a cluster of high-performance clusters. To ease the difficulty of programming with message passing, which is the conventional programming paradigm on cluster systems, many DSM (distributed shared memory) systems have been developed in recent years. However, there have been no DSM systems developed on multi-clusters. The DSM system consists of a runtime system to support basic functions for accessing virtual shared memory built on such environment. The functions are allocation of global data, read and write accesses to global data, synchronization of the whole system, and mutual exclusion. The authors have evaluated the performance of the runtime system, built on a SMP cluster, COMPaS, at RWCP(Real World Computing Partnership) in Tsukuba, Japan. The result shows that a read access to remote memory on the same cluster costs about 0.2msec, while a read access to remote memory on other cluster costs about 1.3msec. The execution time of LU decomposition on a multi-cluster consisting two clusters of three PCs is about 2.8times faster than the time on one PC.",
author = "Takeshi Nanri and Yoshitaka Watanabe and Hiyoyuki Sato and Masaaki Shimasaki",
year = "2000",
month = "12",
day = "1",
language = "English",
isbn = "8489925704",
booktitle = "European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000",

}

TY - GEN

T1 - Preliminary investigation of distributed shared memory system on a cluster of high performance clusters

AU - Nanri, Takeshi

AU - Watanabe, Yoshitaka

AU - Sato, Hiyoyuki

AU - Shimasaki, Masaaki

PY - 2000/12/1

Y1 - 2000/12/1

N2 - This paper introduces design and basic performance of the DSM(distributed shared memory) system on a cluster of clusters. Networking devices such as Myrinet have improved the performance of cluster systems significantly. In addition to that, such kind of network devices introduced a new hierarchical architecture; a multi-cluster, a cluster of high-performance clusters. To ease the difficulty of programming with message passing, which is the conventional programming paradigm on cluster systems, many DSM (distributed shared memory) systems have been developed in recent years. However, there have been no DSM systems developed on multi-clusters. The DSM system consists of a runtime system to support basic functions for accessing virtual shared memory built on such environment. The functions are allocation of global data, read and write accesses to global data, synchronization of the whole system, and mutual exclusion. The authors have evaluated the performance of the runtime system, built on a SMP cluster, COMPaS, at RWCP(Real World Computing Partnership) in Tsukuba, Japan. The result shows that a read access to remote memory on the same cluster costs about 0.2msec, while a read access to remote memory on other cluster costs about 1.3msec. The execution time of LU decomposition on a multi-cluster consisting two clusters of three PCs is about 2.8times faster than the time on one PC.

AB - This paper introduces design and basic performance of the DSM(distributed shared memory) system on a cluster of clusters. Networking devices such as Myrinet have improved the performance of cluster systems significantly. In addition to that, such kind of network devices introduced a new hierarchical architecture; a multi-cluster, a cluster of high-performance clusters. To ease the difficulty of programming with message passing, which is the conventional programming paradigm on cluster systems, many DSM (distributed shared memory) systems have been developed in recent years. However, there have been no DSM systems developed on multi-clusters. The DSM system consists of a runtime system to support basic functions for accessing virtual shared memory built on such environment. The functions are allocation of global data, read and write accesses to global data, synchronization of the whole system, and mutual exclusion. The authors have evaluated the performance of the runtime system, built on a SMP cluster, COMPaS, at RWCP(Real World Computing Partnership) in Tsukuba, Japan. The result shows that a read access to remote memory on the same cluster costs about 0.2msec, while a read access to remote memory on other cluster costs about 1.3msec. The execution time of LU decomposition on a multi-cluster consisting two clusters of three PCs is about 2.8times faster than the time on one PC.

UR - http://www.scopus.com/inward/record.url?scp=84893363678&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893363678&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84893363678

SN - 8489925704

SN - 9788489925700

BT - European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS 2000

ER -