Analysis of the statistical impact of variability in a 12 nm nanosheet FET

New gate-all-around transistor architectures as nanosheet FETs are proposed nowadays to substitute FinFETs, because of their better gate electrostatic control along the conduction channel. To explore these architectures it is necessary to estimate the impact of different sources of variability on their performance. In this work, we present a statistical study based on 3D simulations of a 12 nm gate length (Lg) nanosheet FET suffering from the impact of two different sources of variability: metal grain granularity (MGG), and line edge roughness (LER). Also, we evaluate the correlation between different figures of merit (FoM), namely the threshold voltage (Vth ), the off current (Ioff), the subthreshold slope (SS) and the on current (Ion ).