Fast and accurate estimation of nano-scaled SRAM read failure probability using critical point sampling

Variability in process parameters is making accurate estimate of nano-scale SRAM stability an extremely challenging task. In this paper, we propose a new method to detect the read failure in an SRAM cell using critical point sampling technique. Using this technique, we propose two types of read failure probability estimation method, (1) quasi-analytical and (2) completely analytical method. The result shows that our proposed model can achieve high accuracy, while being 20× faster in computational speed. Our method can be applied to different phases of design to reduce the overall design time, and can be used for optimizing the given design in order to obtain a better yield.