Performance Evaluation of Distributed Memory Multiprocessor Systems
Djamel Benazzouz · Intelligent Automation & Soft Computing · 2001
Abstract This paper presents a modelling approach for the performance evaluation of switching network with a baseline topology. Stochastic behavior is taken into account with randomized requests. The random processes which are presented in this paper are homogenous and finite Markov chains where the state space is discrete and time is continuous. Two strategies of the model were proposed; which are the unbuffered and buffered. It was shown that the unbuffered strategy is better suited for a low blocking probabilities.