VFlow: Discovering Optimal Agentic Workflows for Verilog Generation

Abstract: Hardware design automation faces challenges in generating high-quality Verilog code efficiently. This paper introduces VFlow, an automated framework that optimizes agentic workflows for Verilog code generation. Unlike existing approaches that rely on pre-defined prompting strategies, VFlow leverages Monte Carlo Tree Search (MCTS) to discover effective sequences of LLMs invocations that maximize code quality while minimizing computational costs. VFlow extends the AFLOW methodology with domain-specific operators addressing hardware design requirements, including syntax validation, simulation-based verification, and synthesis optimization. Experimental evaluation on the VerilogEval benchmark demonstrates VFlow's superiority, achieving an 83.6% average pass@1 rate-a 6.1\% improvement over state-of-the-art PromptV and a 36.9\% gain compared to direct LLM invocation. Most significantly, VFlow enhances the capabilities of smaller models, enabling DeepSeek-V3 to achieve 141.2\% of GPT-4o's performance while reducing API costs to just 13\%. These findings indicate that intelligently optimized workflows enable cost-efficient LLMs to outperform larger models on hardware design tasks, potentially democratizing access to advanced digital circuit development tools and accelerating innovation in the semiconductor industry