Machine Learning-Driven Paradigms in VLSI Design: Exploring the Synergy with Pruning for Optimal Circuit Efficiency

This paper investigates the incorporation of machine learning (ML) techniques in Very Large-Scale Integration (VLSI) design, concentrating on pruning methods to boost circuit efficiency. By utilizing ML algorithms for feature extraction, optimization, and decision-making while applying pruning methods to eliminate redundancy, our study introduces a unique strategy for achieving superior circuit performance. Key contributions include a thorough methodology that describes ML- based pruning in hardware design along with an assessment of its effects on design metrics like area, power, and speed. The findings indicate substantial advancements, implying that ML-based pruning could transform VLSI design processes.