ON DEPENDABILITY EVALUATION OF MESH-CONNECTED PROCESSORS

Analytical techniques for reliability and availability prediction of m esh-connected systems are proposed in this paper, The models are based on the submesh requirements, First, a reliability model is proposed a ssuming that a submesh can be always recognized if it exits, Analysis of the linear consecutive n-out-of-N system is extended using an expan ding row/column technique to evaluate the submesh reliability, An alte rnative approach called row folding is also discussed. Due to the high complexity involved in computing the exact reliability, both of these techniques use approximation to estimate lower bounds, Next, the subm esh reliability is computed based on two different allocation policies , known as the two-dimensional buddy system (TDBS), and the frame slid ing (FS). The model with the TDBS is further extended to estimate the reliability of multiple working submeshes, which is useful in a multiu ser environment, Availability analysis for a submesh of the required s ize is conducted using a Markov chain (MC), State truncation is used t o reduce the computation time, and the MC is solved using a software p ackage called HARP, Validation of the analytical models is done throug h extensive simulation, Issues, such as reliability comparison based o n allocation policies, and methods for improving system reliability ar e addressed using the analytical models.