Physics-based modeling of FinFET RF variability under Shorted- and Independent-Gates bias

Abstract

FinFETs operated with varying bias, and in particular with Short-circuited Gates (SG) or Independent Gates (IG), are actively investigated for RF analog applications. The device process variability is known to vary, at least for DC performances, according to the FINFET bias. This paper presents a novel, comprehensive physics-based variability analysis focused on AC parameters for a double-gate (DG) MOSFET (FinFET) both in SG and IG conditions. The analysis is carried out with a numerically efficient Green's Function technique [1], [2], that exploits a nonlinear variability analysis tool in quasi-linear condition. The AC variability analysis of the FinFET includes selected geometrical and physical parameters, such as the fin width, the source/drain-gate distance and the doping level, whose role is especially relevant for the extraction of the device parasitics' variations. We demonstrate that the sensitivity of the AC parameters differs in the IG and SG case, especially concerning gate capacitances.