Electrical-Performance Characteristics Prediction of Gate All Around Tunnel FET Using Machine Learning

In this article, the nano-wire gate-all-around tunnel FET (GAA-NWTFET) has been designed and its performance characteristics have been predicted using technology computer-aided design-extended machine learning (TCAD-ML). The voltage-current (V-I) curves for input and output characteristics, in forward voltage sweeps, have been predicted with low mean absolute error, mean square error, and root mean square error by using the K nearest neighbor algorithm. Moreover, transfer and output characteristics along with the transconductance and capacitance have been predicted accurately using TCAD-ML. This investigation confirmed that the computational time for device development can be considerably minimized by using TCAD-ML over conventional TCAD-based simulations.