Artificial intelligence for post-moore field-effect transistors: a review and perspective

Abstract

As Moore’s Law approaches its fundamental physical limits, the development of post-Moore field-effect transistors (FETs) has emerged as a critical pathway to sustain the advancement of semiconductor technology. Artificial intelligence (AI), with its unparalleled capabilities in data-driven modeling, optimization, and predictive analytics, is revolutionizing the design, simulation, and fabrication of next-generation FETs. This review systematically examines the methodological frameworks, algorithmic strategies, and multidimensional applications of AI in FETs research, with particular emphasis on high-throughput screening and performance prediction of channel materials, gate dielectrics, and electrode materials, alongside device architecture optimization. Furthermore, we highlight future opportunities at the intersection of AI to redefine the frontiers of post-Moore electronics. This review aims to inspire multidisciplinary efforts toward AI-empowered FET innovation, bridging the gap between computational intelligence and semiconductor engineering for sustainable technological progress.