An Accurate Analytical SNM Modeling Technique for SRAMs Based on Butterworth Filter Function

With the impact of process variations becoming increasingly pronounced in the sub-90nm regime, the stability of SRAM cells is considerably degraded. Static noise margin (SNM) is popularly used for SRAM stability measurement. Therefore, efficient techniques to estimate SNM are required for robust SRAM design. In this paper, we have proposed an analytical SRAM SNM modeling technique based on Butterworth function and coordinate transformation. The accuracy of the SNM model has been verified by changing of the SRAM design parameters, such as the relative sizes of the 6 transistors and device doping. With a wide range of design parameter change, less than 8% of estimation error is achieved by the proposed SNM model. The proposed SNM model has been applied to analyze the impact of process variations on SRAM SNM distribution. Also, the distribution of read SNM has been estimated for a range of design parameter variations