Advancements in Electronic Design Automation with Large Language Models

The field of Electronic Design Automation (EDA) is witnessing a significant shift with the integration of Large Language Models (LLMs). Recent developments have demonstrated the potential of LLMs in simplifying and automating various aspects of the design-to-manufacturing workflow. Notably, LLMs are being leveraged to generate high-quality code, automate design tasks, and enhance verification and testing processes. The use of reinforcement learning and multi-agent architectures is also being explored to improve the performance and efficiency of LLMs in EDA. These advancements have the potential to significantly accelerate hardware development and reduce the complexity of modern integrated circuits. Noteworthy papers in this area include ASIC-Agent, which presents an autonomous multi-agent system for ASIC design, and VERIRL, which introduces a reinforcement learning framework for Verilog code generation. GENIE-ASI is also notable for its training-free, LLM-based methodology for analog subcircuit identification.

Sources

ASIC-Agent: An Autonomous Multi-Agent System for ASIC Design with Benchmark Evaluation

VERIRL: Boosting the LLM-based Verilog Code Generation via Reinforcement Learning

GENIE-ASI: Generative Instruction and Executable Code for Analog Subcircuit Identification

Large Language Models (LLMs) for Electronic Design Automation (EDA)

Exploiting Instantiations from Paramodulation Proofs in Isabelle/HOL

Built with on top of