Multi-Objective Coevolution of Prompts and Templates for Circuit Approximation
For hardware designers, this work automates the design of approximate multipliers with better trade-offs, though it is incremental as it applies existing co-evolution and LLM techniques to a specific domain.
The paper introduces a co-evolutionary algorithm using an off-the-shelf LLM to automate the design of 8-bit approximate multipliers, achieving improved error-area trade-offs compared to optimized circuits from EvoApproxLib.
Approximate multipliers deliberately relax computational accuracy to achieve gains in power efficiency, latency, and silicon area, which makes them well-suited for error-resilient applications such as neural networks. In this work, we introduce a co-evolutionary algorithm that leverages an off-the-shelf large language model (LLM) without requiring domain-specific training to automate the design of optimized 8-bit approximate multipliers. The approach simultaneously evolves a population of candidate circuits and a population of prompt templates that steer LLM-driven modifications. Experimental results for several target design objectives demonstrate that the proposed method discovers approximate multipliers with improved error-area trade-offs compared to highly optimized circuits from the EvoApproxLib library.