CRADLE: Conversational RTL Design Space Exploration with LLM-based Multi-Agent Systems

2025-08-12

AI Summary

Traditional RTL design space exploration suffers from rigidity, limited user interactivity, and insufficient adaptability for optimization. To address these limitations, this paper proposes a large language model (LLM)-based multi-agent conversational framework. It integrates generative and evaluative agents that collaboratively enable user-guided dynamic RTL generation, autonomous verification, error correction, and iterative optimization—driven by FPGA resource minimization (e.g., LUTs and FFs). Unlike static pipeline approaches, our work pioneers the adoption of conversational interaction for RTL design space exploration, establishing an end-to-end semantic-level feedback loop. Experimental evaluation on the RTLLM benchmark demonstrates that the proposed method achieves average reductions of 48% in LUT usage and 40% in flip-flop (FF) consumption, significantly enhancing hardware efficiency and design flexibility.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

This paper presents CRADLE, a conversational framework for design space exploration of RTL designs using LLM-based multi-agent systems. Unlike existing rigid approaches, CRADLE enables user-guided flows with internal self-verification, correction, and optimization. We demonstrate the framework with a generator-critic agent system targeting FPGA resource minimization using state-of-the-art LLMs. Experimental results on the RTLLM benchmark show that CRADLE achieves significant reductions in resource usage with averages of 48% and 40% in LUTs and FFs across all benchmark designs.

Problem

• Enables conversational RTL design space exploration
• Provides self-verification and optimization for RTL designs
• Reduces FPGA resource usage significantly

Innovation

• LLM-based multi-agent conversational framework
• Self-verification and correction mechanisms
• FPGA resource optimization via generator-critic system