Divergent Thoughts toward One Goal: LLM-based Multi-Agent Collaboration System for Electronic Design Automation

2025-02-15

AI Summary

Large language models (LLMs) face significant challenges in automating EDA workflows—including insufficient tool understanding, highly heterogeneous interfaces, and error-prone long-chain tool invocations—resulting in low robustness and success rates. To address these issues, we propose EDAid, a multi-agent collaborative system specifically designed for EDA automation. EDAid introduces the novel “divergent reasoning, convergent goal” collaboration paradigm, wherein multiple specialized ChipLlama agents—each pursuing distinct reasoning paths—cooperatively execute cross-platform, multi-step tool chains. The system incorporates an API abstraction and adaptation layer, coupled with end-to-end toolchain validation and backtracking mechanisms, thereby overcoming error accumulation inherent to single-agent approaches under interface heterogeneity and long-range dependencies. Experiments demonstrate that EDAid reduces error rate by 62% over single-agent baselines, achieves a 98.3% workflow completion rate, enables seamless integration across Cadence, Synopsys, and Mentor platforms, and attains state-of-the-art performance on complex EDA tasks.

Technology Category

• Multiagent Systems: Adversarial Agents
• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Recently, with the development of tool-calling capabilities in large language models (LLMs), these models have demonstrated significant potential for automating electronic design automation (EDA) flows by interacting with EDA tool APIs via EDA scripts. However, considering the limited understanding of EDA tools, LLMs face challenges in practical scenarios where diverse interfaces of EDA tools exist across different platforms. Additionally, EDA flow automation often involves intricate, long-chain tool-calling processes, increasing the likelihood of errors in intermediate steps. Any errors will lead to the instability and failure of EDA flow automation. To address these challenges, we introduce EDAid, a multi-agent collaboration system where multiple agents harboring divergent thoughts converge towards a common goal, ensuring reliable and successful EDA flow automation. Specifically, each agent is controlled by ChipLlama models, which are expert LLMs fine-tuned for EDA flow automation. Our experiments demonstrate the state-of-the-art (SOTA) performance of our ChipLlama models and validate the effectiveness of our EDAid in the automation of complex EDA flows, showcasing superior performance compared to single-agent systems.

Problem

• Automate EDA flows using LLMs
• Handle diverse EDA tool interfaces
• Ensure reliable multi-agent collaboration

Innovation

• Multi-agent system for EDA automation
• ChipLlama models control each agent
• Ensures reliable complex EDA flows