Analytical Estimation of Threshold Voltage Variability by Metal Gate Granularity in FinFET

Abstract

Metal gate granularity (MGG)-induced threshold voltage variability is the dominant source of variability in FinFETs. The analytical model for MGG-based variability is essential to study its circuit impact. In this paper, we present a novel electrostatics and percolation theory-based analytical model to estimate MGG-induced threshold voltage (VT) variability. The model is capable of analyzing realistic grain shapeand position-distributions-demonstrated with random Voronoi grains. The model is benchmarked against stochastic 3-D TCAD simulations to demonstrate excellent accuracy [4% error in σ (VT)]. Furthermore, it shows in excess of 12× improvement in accuracy over existing analytical models. Our model enables a 55× reduction in the computation time in comparison with 3-D stochastic TCAD simulations.