Applying Existing Large Language Models for Print Circuit Board Routing

Large language models (LLMs), such as GPT-4.0 and Gemini, have achieved excellent performance on natural-language tasks, and they also show high expectations for logical reasoning. In the realm of print circuit board (PCB) routing, complex routing scenarios still rely on manual routing performed by seasoned engineers, which consumes significant human resources and time. This paper proposes an approach using few-shot and chain-of-thought training LLMs to tackle this issue, enabling LLMs to assist engineers in design tasks with a small number of samples. We tested the performance of LLMs in different routing scenarios with a few examples, validating the applicability of this method. Furthermore, we explored fine-tuning techniques to enhance the effectiveness of the few-shot learning approach, to overcome the limitation of scarce real-world PCB cases, and we employed code synthetic cases to fine-tune the model in place of actual PCB scenarios, ultimately improving the LLMs’ capability to manage intricate routing tasks. The results validate the feasibility and effectiveness of this method, offering a promising avenue for reducing the manual burden in PCB design.