Compact Modeling of Process Variation and Reliability Predictions for Nanosheet Gate-All-Around FET

In this work, a semi-analytical compact model is developed to quantify the impact of random process variations on nanosheet field-effect transistors (NSFETs) at the 3nm technology node. Three primary sources of variability work function variation (WFV), line width roughness (LWR), and gate edge roughness (GER) are systematically analyzed. By extracting and calibrating empirical parameters, the proposed model accurately captures the statistical trends of process-induced fluctuations across a broad range of conditions. The model is integrated into the BSIM-CMG framework for circuit-level variability assessment, enabling comprehensive evaluation of performance deviations. Simulation results indicate that WFV dominates the overall reliability degradation, leading to energy variations from -12% to +24%. This study provides a refined predictive framework for assessing process-induced reliability risks and optimizing circuit design in advanced semiconductor technologies.