Spatial-correlation-aware soft error rate analysis using quasi-importance sampling

Since statistical methods are important to accurately estimate the soft error rate (SER) of circuits with process variations, we incorporate the spatial correlation into SSER analysis to provide better accuracy. Moreover, the SSER analysis based on quasi-Monte Carlo comes into the difficulty of sampling points on a non-uniform distribution or unbounded distribution. Therefore, in this paper, we employ the quasi-importance sampling into Monte-Carlo simulation to overcome such sampling issue. Experimental results show that the quasi-importance sampling Monte-Carlo SSER analysis framework is capable of more precisely estimating circuit SSERs and reaches 3.72X speedups when compared to the baseline Monte-Carlo simulation.