By Topic

Interference Analysis of Shuffle/Exchange Networks

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Thanawastien, S. ; Department of Electrical Engineering, Auburn University ; Nelson, V.P.

The use of shuffle/exchange (S/E) interconnection networks in multiprocessor systems has been proposed for several applications. In order to evaluate the potential performance and reliability of such systems, the effects of conflicts involving switch and memory contention should be determined. This paper presents a discrete Markov chain model to study the effects of such contention for S/E networks used in random access application environments. This model is used to derive memory bandwidth (MBW) of the system, with analytic expressions presented for 4 X 4 and 8 X 8 S/E networks. Two cases are considered. First, the model is applied to systems in which all processors issue their access requests synchronously, then this model is used to construct a traffic model which allows the generation of access requests at arbitrary times. The results of the analysis are compared with simulation results and the analytical results derived by Bhandarkar for the full Crossbar network. Besides the MBW, other interference-related parameters such as blocking probability, traffic, loading, and blocking delay are discussed. In particular, the loading effect of a 16 X 16 S/E network obtained from the model is compared with the results of the simulation from Wirsching's CONET model.

Published in:

Computers, IEEE Transactions on  (Volume:C-30 ,  Issue: 8 )