Performance tuning of adaptive wormhole routing through selection function choice

  • Authors:
  • Loren Schwiebert;Renelius Bell

  • Affiliations:
  • Department of Computer Science, Wayne State University, Detroit, MI;Department of Electrical and Computer Engineering, Wayne State University, Detroit, Michigan

  • Venue:
  • Journal of Parallel and Distributed Computing
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

Many adaptive routing algorithms have been proposed for wormhole-routed interconnection networks. Comparatively little work, however, has been done on determining how the selection function (routing policy) affects the performance of an adaptive routing algorithm. In this paper, we present a detailed simulation study of various selection functions for fully adaptive wormhole routing on two-dimensional meshes. The simulation results show that the choice of selection function has a significant effect on the average message latency. In addition, it is possible to find a single selection function that exhibits excellent performance across a wide range of traffic patterns, network sizes, and number of virtual channels. Thus, a well-chosen selection function for an adaptive routing algorithm can lead to consistently better performance than an arbitrary selection function. One of the selection functions considered is a theoretically optimal selection function (IEEE Trans. Comput. 38(10) (October 1989), 1362-1371). We show that although theoretically optimal, the actual performance of the optimal selection function is not always best. An explanation and interpretation of the results is provided.