An approach based on neural computation to simulate the nanoscale CMOS circuits: Application to the simulation of CMOS inverter

Abstract

As critical transistor dimensions scale below the 100nm (nanoscale) regime, quantum mechanical (QM) effects begin to manifest themselves and affect important device performance metrics. Therefore, simulation tools which can be applied to design nanoscale transistors in the future, require new theory and modeling techniques that capture the physics of quantum transport accurately and efficiently. In this paper, we apply an artificial neural network (ANN) to the study of the nanoscale CMOS circuits. The latter is based on the 2-D numerical non-equilibrium Green’s function (NEGF) simulation of the current–voltage characteristics of an undoped symmetric DG MOSFET. The encouraging comparisons between numerical results and ANN PSPICE simulations have indicated that the developed ANN subcircuit representation particularly suitable to be incorporated in SPICE-like tools for nanoscale CMOS circuits simulation.