On the design of scalable pipelined broadcasting for mesh networks

Ahmed Y. Al-Dubai, Mohamed Ould-Khaoua

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

Abstract

Minimising the communication latency and achieving considerable scalability are of paramount importance when designing high performance broadcast algorithms. Many algorithms for wormhole-switched meshes have been widely reported in the literature. However, most of these algorithms handle broadcast in a sequential manner and do not scale well with the network size. As a consequence, many parallel applications cannot be efficiently supported using existing algorithms. Motivated by these observations, this paper presents a new broadcast algorithm for the all-port mesh networks. The unique feature of the proposed algorithm is its capability of handling broadcast in only one message-passing step irrespective of the network size. Results from a comparative analysis and simulation reveal that the proposed algorithm exhibits superior performance characteristics over those of the well-known Recursive Doubling, Extending Dominating Node and Network Partitioning algorithms.

Original languageEnglish
Title of host publicationProceedings - International Symposium on High Performance Computing Systems and Applications
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages98-105
Number of pages8
Volume2002-January
ISBN (Print)0769516262
DOIs
Publication statusPublished - 2002
Event16th Annual International Symposium on High Performance Computing Systems and Applications, HPCS 2002 - Moncton, Canada
Duration: Jun 16 2002Jun 19 2002

Other

Other16th Annual International Symposium on High Performance Computing Systems and Applications, HPCS 2002
CountryCanada
CityMoncton
Period6/16/026/19/02

Fingerprint

Broadcasting
Message passing
Scalability
Communication

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Al-Dubai, A. Y., & Ould-Khaoua, M. (2002). On the design of scalable pipelined broadcasting for mesh networks. In Proceedings - International Symposium on High Performance Computing Systems and Applications (Vol. 2002-January, pp. 98-105). [1019140] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/HPCSA.2002.1019140

On the design of scalable pipelined broadcasting for mesh networks. / Al-Dubai, Ahmed Y.; Ould-Khaoua, Mohamed.

Proceedings - International Symposium on High Performance Computing Systems and Applications. Vol. 2002-January Institute of Electrical and Electronics Engineers Inc., 2002. p. 98-105 1019140.

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

Al-Dubai, AY & Ould-Khaoua, M 2002, On the design of scalable pipelined broadcasting for mesh networks. in Proceedings - International Symposium on High Performance Computing Systems and Applications. vol. 2002-January, 1019140, Institute of Electrical and Electronics Engineers Inc., pp. 98-105, 16th Annual International Symposium on High Performance Computing Systems and Applications, HPCS 2002, Moncton, Canada, 6/16/02. https://doi.org/10.1109/HPCSA.2002.1019140
Al-Dubai AY, Ould-Khaoua M. On the design of scalable pipelined broadcasting for mesh networks. In Proceedings - International Symposium on High Performance Computing Systems and Applications. Vol. 2002-January. Institute of Electrical and Electronics Engineers Inc. 2002. p. 98-105. 1019140 https://doi.org/10.1109/HPCSA.2002.1019140
Al-Dubai, Ahmed Y. ; Ould-Khaoua, Mohamed. / On the design of scalable pipelined broadcasting for mesh networks. Proceedings - International Symposium on High Performance Computing Systems and Applications. Vol. 2002-January Institute of Electrical and Electronics Engineers Inc., 2002. pp. 98-105
@inproceedings{d27c6bb559234d50a53eb6114b832034,
title = "On the design of scalable pipelined broadcasting for mesh networks",
abstract = "Minimising the communication latency and achieving considerable scalability are of paramount importance when designing high performance broadcast algorithms. Many algorithms for wormhole-switched meshes have been widely reported in the literature. However, most of these algorithms handle broadcast in a sequential manner and do not scale well with the network size. As a consequence, many parallel applications cannot be efficiently supported using existing algorithms. Motivated by these observations, this paper presents a new broadcast algorithm for the all-port mesh networks. The unique feature of the proposed algorithm is its capability of handling broadcast in only one message-passing step irrespective of the network size. Results from a comparative analysis and simulation reveal that the proposed algorithm exhibits superior performance characteristics over those of the well-known Recursive Doubling, Extending Dominating Node and Network Partitioning algorithms.",
author = "Al-Dubai, {Ahmed Y.} and Mohamed Ould-Khaoua",
year = "2002",
doi = "10.1109/HPCSA.2002.1019140",
language = "English",
isbn = "0769516262",
volume = "2002-January",
pages = "98--105",
booktitle = "Proceedings - International Symposium on High Performance Computing Systems and Applications",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - On the design of scalable pipelined broadcasting for mesh networks

AU - Al-Dubai, Ahmed Y.

AU - Ould-Khaoua, Mohamed

PY - 2002

Y1 - 2002

N2 - Minimising the communication latency and achieving considerable scalability are of paramount importance when designing high performance broadcast algorithms. Many algorithms for wormhole-switched meshes have been widely reported in the literature. However, most of these algorithms handle broadcast in a sequential manner and do not scale well with the network size. As a consequence, many parallel applications cannot be efficiently supported using existing algorithms. Motivated by these observations, this paper presents a new broadcast algorithm for the all-port mesh networks. The unique feature of the proposed algorithm is its capability of handling broadcast in only one message-passing step irrespective of the network size. Results from a comparative analysis and simulation reveal that the proposed algorithm exhibits superior performance characteristics over those of the well-known Recursive Doubling, Extending Dominating Node and Network Partitioning algorithms.

AB - Minimising the communication latency and achieving considerable scalability are of paramount importance when designing high performance broadcast algorithms. Many algorithms for wormhole-switched meshes have been widely reported in the literature. However, most of these algorithms handle broadcast in a sequential manner and do not scale well with the network size. As a consequence, many parallel applications cannot be efficiently supported using existing algorithms. Motivated by these observations, this paper presents a new broadcast algorithm for the all-port mesh networks. The unique feature of the proposed algorithm is its capability of handling broadcast in only one message-passing step irrespective of the network size. Results from a comparative analysis and simulation reveal that the proposed algorithm exhibits superior performance characteristics over those of the well-known Recursive Doubling, Extending Dominating Node and Network Partitioning algorithms.

UR - http://www.scopus.com/inward/record.url?scp=84949908118&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84949908118&partnerID=8YFLogxK

U2 - 10.1109/HPCSA.2002.1019140

DO - 10.1109/HPCSA.2002.1019140

M3 - Conference contribution

AN - SCOPUS:84949908118

SN - 0769516262

VL - 2002-January

SP - 98

EP - 105

BT - Proceedings - International Symposium on High Performance Computing Systems and Applications

PB - Institute of Electrical and Electronics Engineers Inc.

ER -