Structural Analysis of Multi-Core Processor and Reliability Evaluation Model
S. Tsiramua, H. Meladze, T. Davitashvili, J. M. Sanchez, F. Criado-Aldeanueva
TL;DR
The paper addresses reliability challenges in large-scale, variable-structure multi-core processors by developing a logical-probabilistic framework that treats cores as multifunctional elements (MFEs). It introduces state-space models for MFEs, a reconfigurable-system reliability formulation using a 0-1 resource matrix and matrix permanents to quantify flexibility, and an orthogonalization-based method to convert Boolean operability into probabilistic estimates, yielding reliability polynomials. Through comparative analysis of 2-core and 4-core configurations, the authors demonstrate that added cores improve flexibility, reliability, fault tolerance, and structure perfection, validating the approach with concrete results and polynomials. The methods offer a principled way to assess and optimize reconfigurable multi-core designs, with future work pointing to dynamic time-aware models such as Markov processes for richer realism.
Abstract
In the present paper, the models of structural analysis and evaluation of efficiency indicators (reliability, fault tolerance, viability, and flexibility) of a multi core processor with variable structure, equipped with multi functional cores, are considered. Using logical probabilistic methods, the following has been developed: models for evaluating the reliability and fault tolerance of processor cores as multi functional elements; logical probabilistic models of the shortest paths, flexibility, and performance conditions for successful operation of multi core processors based on multi functional cores; and models for estimating the reliability, fault tolerance, and lifetime of multi core processors considering all possible states of performance. The results of the structural analysis of two core and four core processors and the trends of increasing the efficiency indicators of multi core processors are presented.