University of Hertfordshire

By the same authors

An adaptive, reconfigurable interconnect for computational clusters

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

Documents

  • 902273

    Accepted author manuscript, 86 KB, PDF document

View graph of relations
Original languageEnglish
Title of host publicationProcs 1st IEEE/ACM Int Symposium on Cluster Computing & the Grid
Pages229-236
Publication statusPublished - 2001

Abstract

This paper describes the principles of an original adaptive interconnect for a computational cluster. Torus topology (2d or 3d) is used as a basis but nodes are allowed to effectively migrate along the torus cycles. An optoelectronic scheme which makes such migrations possible with only local synchronisation is outlined. Between the instances of migration the interconnect behaves as a direct packet-routing network which constantly monitors its traffic parameters. A decentralised predictive algorithm is applied periodically to decide whether the current topology is consistent with the predominant traffic flow and if it is not, a reconfiguration to a better-matched topology occurs. We present simulation results that show that on some standard computational benchmarks a significant speedup is possible as a result of automatic matching between the effective topology of the application’s message-passing infrastructure and that of the interconnect.

ID: 462012