AI/ML Assisted VLSI Design Automation

Abstract

ABSTRACT The VLSI (Very Large Scale Integration) design process has become increasingly complex due to the scaling of technology nodes and the rising demand for high-performance, low-power integrated circuits. Traditional Electronic Design Automation (EDA) tools often face challenges in achieving optimal design solutions efficiently. Artificial Intelligence (AI) and Machine Learning (ML) techniques are emerging as promising approaches to enhance VLSI design automation by improving tasks such as logic synthesis, placement, routing, and verification. This review paper discusses recent advancements in AI/ML-assisted VLSI design, highlighting the applications of supervised learning, reinforcement learning, and deep learning in different stages of the design flow. Various case studies and comparative analyses of AI-driven methodologies versus conventional techniques are presented. The paper also explores future directions and challenges in integrating AI/ML into mainstream VLSI design workflows.

KEYWORDS: VLSI, AI, ML, EDA, Placement, Routing, Logic Synthesis, Verification, Neural Networks, Reinforcement Learning