`Express channel' concept in hypermeshes and k-ary n-cubes

S. Loucif, L. M. Mackenzie, M. Ould-Khaoua

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

3 Citations (Scopus)

Abstract

Low-dimensional k-ary n-cubes have been popular in recent multicomputers. However, these networks suffer from high switching delays due to their high message distance. To overcome this problem, Dally has proposed express k-ary n-cubes with express channels, that allow non-local messages to partially bypass clusters of nodes within a dimension. This paper argues that hypergraph topologies, that provide total bypasses within a dimension, represent potential candidates as future high-performance networks. It presents a comparative study, of a regular hypergraph, referred to as the Distributed Crossbar Switch Hypermesh (DCSH), and the express k-ary n-cube, taking into account channel bandwidth constraints which apply in VLSI and multiple-chip technology. The study concludes that the DCSH's total bypass strategy yields superior performance characteristics to the partial bypassing of its express cube counterpart.

Original languageEnglish
Title of host publicationIEEE Symposium on Parallel and Distributed Processing - Proceedings
Editors Anon
PublisherIEEE
Pages566-569
Number of pages4
Publication statusPublished - 1996
EventProceedings of the 1996 8th IEEE Symposium on Parallel and Distributed Processing - New Orleans, LA, USA
Duration: Oct 23 1996Oct 26 1996

Other

OtherProceedings of the 1996 8th IEEE Symposium on Parallel and Distributed Processing
CityNew Orleans, LA, USA
Period10/23/9610/26/96

Fingerprint

Network performance
Switches
Topology
Bandwidth

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Loucif, S., Mackenzie, L. M., & Ould-Khaoua, M. (1996). `Express channel' concept in hypermeshes and k-ary n-cubes. In Anon (Ed.), IEEE Symposium on Parallel and Distributed Processing - Proceedings (pp. 566-569). IEEE.

`Express channel' concept in hypermeshes and k-ary n-cubes. / Loucif, S.; Mackenzie, L. M.; Ould-Khaoua, M.

IEEE Symposium on Parallel and Distributed Processing - Proceedings. ed. / Anon. IEEE, 1996. p. 566-569.

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

Loucif, S, Mackenzie, LM & Ould-Khaoua, M 1996, `Express channel' concept in hypermeshes and k-ary n-cubes. in Anon (ed.), IEEE Symposium on Parallel and Distributed Processing - Proceedings. IEEE, pp. 566-569, Proceedings of the 1996 8th IEEE Symposium on Parallel and Distributed Processing, New Orleans, LA, USA, 10/23/96.
Loucif S, Mackenzie LM, Ould-Khaoua M. `Express channel' concept in hypermeshes and k-ary n-cubes. In Anon, editor, IEEE Symposium on Parallel and Distributed Processing - Proceedings. IEEE. 1996. p. 566-569
Loucif, S. ; Mackenzie, L. M. ; Ould-Khaoua, M. / `Express channel' concept in hypermeshes and k-ary n-cubes. IEEE Symposium on Parallel and Distributed Processing - Proceedings. editor / Anon. IEEE, 1996. pp. 566-569
@inproceedings{b92c16eae3164629b0c16376f89d4d76,
title = "`Express channel' concept in hypermeshes and k-ary n-cubes",
abstract = "Low-dimensional k-ary n-cubes have been popular in recent multicomputers. However, these networks suffer from high switching delays due to their high message distance. To overcome this problem, Dally has proposed express k-ary n-cubes with express channels, that allow non-local messages to partially bypass clusters of nodes within a dimension. This paper argues that hypergraph topologies, that provide total bypasses within a dimension, represent potential candidates as future high-performance networks. It presents a comparative study, of a regular hypergraph, referred to as the Distributed Crossbar Switch Hypermesh (DCSH), and the express k-ary n-cube, taking into account channel bandwidth constraints which apply in VLSI and multiple-chip technology. The study concludes that the DCSH's total bypass strategy yields superior performance characteristics to the partial bypassing of its express cube counterpart.",
author = "S. Loucif and Mackenzie, {L. M.} and M. Ould-Khaoua",
year = "1996",
language = "English",
pages = "566--569",
editor = "Anon",
booktitle = "IEEE Symposium on Parallel and Distributed Processing - Proceedings",
publisher = "IEEE",

}

TY - GEN

T1 - `Express channel' concept in hypermeshes and k-ary n-cubes

AU - Loucif, S.

AU - Mackenzie, L. M.

AU - Ould-Khaoua, M.

PY - 1996

Y1 - 1996

N2 - Low-dimensional k-ary n-cubes have been popular in recent multicomputers. However, these networks suffer from high switching delays due to their high message distance. To overcome this problem, Dally has proposed express k-ary n-cubes with express channels, that allow non-local messages to partially bypass clusters of nodes within a dimension. This paper argues that hypergraph topologies, that provide total bypasses within a dimension, represent potential candidates as future high-performance networks. It presents a comparative study, of a regular hypergraph, referred to as the Distributed Crossbar Switch Hypermesh (DCSH), and the express k-ary n-cube, taking into account channel bandwidth constraints which apply in VLSI and multiple-chip technology. The study concludes that the DCSH's total bypass strategy yields superior performance characteristics to the partial bypassing of its express cube counterpart.

AB - Low-dimensional k-ary n-cubes have been popular in recent multicomputers. However, these networks suffer from high switching delays due to their high message distance. To overcome this problem, Dally has proposed express k-ary n-cubes with express channels, that allow non-local messages to partially bypass clusters of nodes within a dimension. This paper argues that hypergraph topologies, that provide total bypasses within a dimension, represent potential candidates as future high-performance networks. It presents a comparative study, of a regular hypergraph, referred to as the Distributed Crossbar Switch Hypermesh (DCSH), and the express k-ary n-cube, taking into account channel bandwidth constraints which apply in VLSI and multiple-chip technology. The study concludes that the DCSH's total bypass strategy yields superior performance characteristics to the partial bypassing of its express cube counterpart.

UR - http://www.scopus.com/inward/record.url?scp=0030380951&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030380951&partnerID=8YFLogxK

M3 - Conference contribution

SP - 566

EP - 569

BT - IEEE Symposium on Parallel and Distributed Processing - Proceedings

A2 - Anon, null

PB - IEEE

ER -