CROP: Circuit Retrieval and Optimization with Parameter Guidance using LLMs

TL;DR

CROP leverages large language models and embedding-based retrieval to automate and optimize VLSI design parameter tuning, significantly improving efficiency and power consumption in industrial circuit designs.

Contribution

This paper introduces CROP, the first LLM-powered framework for automatic VLSI design flow tuning using semantic retrieval and guided parameter search.

Findings

Achieves 9.9% reduction in power consumption.

Requires fewer iterations than existing methods.

Demonstrates superior quality-of-results on industrial designs.

Abstract

Modern very large-scale integration (VLSI) design requires the implementation of integrated circuits using electronic design automation (EDA) tools. Due to the complexity of EDA algorithms, the vast parameter space poses a huge challenge to chip design optimization, as the combination of even moderate numbers of parameters creates an enormous solution space to explore. Manual parameter selection remains industrial practice despite being excessively laborious and limited by expert experience. To address this issue, we present CROP, the first large language model (LLM)-powered automatic VLSI design flow tuning framework. Our approach includes: (1) a scalable methodology for transforming RTL source code into dense vector representations, (2) an embedding-based retrieval system for matching designs with semantically similar circuits, and (3) a retrieval-augmented generation (RAG)-enhanced LLM-guided parameter search system that constrains the search process with prior knowledge from similar designs. Experiment results demonstrate CROP's ability to achieve superior quality-of-results (QoR) with fewer iterations than existing approaches on industrial designs, including a 9.9% reduction in power consumption.