A Comprehensive Extraction Method to Decompose Different Random Variation Impacts in bulk FinFET Technology

Abstract

A comprehensive variation extraction flow is proposed to separate different random variation sources in FinFET technology. Based on the understanding of the impacts of random dopant fluctuation (RDF), metal gate granularity (MGG), and line edge roughness (LER) on device figures of merit: threshold voltage (V th ) and subthreshold swing (SS), the LER component induced V th variation can be extracted using covariance matrix of SS and V th . The RDF and MGG can be separated from IV curves of FinFETs with different doping concentration. Results decomposed from the overall flow show great agreement with TCAD simulations. The proposed flow can achieve the accurate extraction of the three major variation sources in FinFET technology. Furthermore, the extendibility of the proposed method to understand the physical mechanisms of random variations is exhibited.