An accurate mathematical performance model of partially adaptive routing in binary n-cube multiprocessors

Abstract

Although several analytical models have been proposed in the literature for different interconnection networks with different routing algorithms, there is only one work [M. Ould-Khaoua, An approximate performance model for partially adaptive routing algorithm in hypercubes, Microprocessors and Microsystems 23 (1999) 185–190] dealing with partially adaptive routing. It is not accurate enough especially in heavy traffic regions due to the rough approximation of assuming equal traffic rate over network channels. We show, however, such an approximation can greatly affect the accuracy of the model. This paper proposes an accurate analytical model to predict the average message latency in wormhole hypercube networks using partially adaptive routing. Our model accurately formulates the traffic rate over individual network channels and the results obtained from simulation experiments confirm that the proposed model exhibits a significant accuracy for various network sizes and under different operating conditions even in moderate to heavy traffic regions, the regions for which the model in [M. Ould-Khaoua, An approximate performance model for partially adaptive routing algorithm in hypercubes, Microprocessors and Microsystems 23 (1999) 185–190] fails to exhibit acceptable accuracy.