A Mismatch Model for Random Gate Length Induced Variations in Nanometer MOSFETs

With the CMOS technology entering nanometer regime, random channel length variation has become one of the most important factors which affect the performance and yield of integrated circuits. This paper analyzes and verifies the correlation and mismatch analytical formula for current gain factor and threshold voltage due to this process variation in nanometer MOS devices. And thus a 22 nm MOSFET current mismatch model is proposed according to the propagation of variation methodology. The experimental results show that, the proposed model reveals that short channel effect leads to threshold voltage shift and changes the effective carrier mobility, thereby affects the current gain factor and device currents. The proposed model accurately captures the statistical properties for random gate induced electrical parameter variations.