Emerging VLSI Technologies for High Performance AI and ML Applications: Survey Paper

Modern market technology evolution requires CMOS-based semiconductor manufacturing to use more effective and smarter Electronic Design Automation (EDA) methods because of rising efficiency requirements. This paper examines how cutting-edge technologies consisting of machine learning (ML), artificial intelligence (AI), edge computing and neuromorphic systems function in Very Large Scale Integration (VLSI) and embedded system designs. The emphasis on sustainability happens through "design-based equivalent scaling" as well as AI implementations in chip production and power improvement with AVS alongside in-situ detection and task-memory scheduling. The paper details how FPGAs and MPSoCs bring performance benefits to hardware systems while examining new memory solutions consisting of resistive RAM and in-memory computing technology that help bypass traditional von Neumann system constraints. The joint optimization between hardware and software technology leads to meaningful applications which detect ASD while improving biomedical imaging. The paper demonstrates through diverse academic studies that ML models with energy-efficient circuit designs and edge AI represent future semiconductor and embedded systems standards.