发表机构
Columbia University; Google DeepMind; Google（哥伦比亚大学; 谷歌DeepMind; 谷歌）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对芯片验证中EDA流程仍依赖人工的问题，提出端到端智能体框架BTTF，将仿真数据转为SQLite数据库并配合多智能体编排，实现自然语言查询转SQL，在150查询基准上达到95.33%准确率，推动自主验证。
AI 中文摘要
现代人工智能前所未有的计算规模依赖于复杂的、拥有数十亿晶体管的片上系统，然而验证这些芯片的工作流程仍然顽固地依赖人工。尽管大型语言模型（LLMs）已迅速进入电子设计自动化（EDA）领域，但现有研究中约有74.6%针对静态寄存器传输级（RTL）代码生成，而仿真后验证和交互式波形调试则基本未被触及。我们提出了Back-to-the-Future（BTTF），一个端到端的智能体框架，以弥补这一基础设施缺口。BTTF将海量非结构化的仿真转储提炼为规范化的关系型SQLite数据库，并将其与一个协作式多智能体编排引擎相结合，该引擎能将自然语言验证查询转换为感知模式的SQL，同时将信号异常与版本化的RTL仓库相关联。在包含150个查询的基准测试中，BTTF达到了95.33%的执行准确率，为自主EDA验证指明了一条切实可行的路径。
英文摘要
The unprecedented computational scale of modern artificial intelligence depends on complex, multi-billion-transistor Systems-on-Chip, yet the workflows that verify these chips remain stubbornly manual. Although Large Language Models (LLMs) have made rapid inroads into Electronic Design Automation (EDA), approximately 74.6% of existing studies target static Register-Transfer Level (RTL) code generation, leaving post-simulation verification and interactive waveform debugging largely untouched. We introduce Back-to-the-Future (BTTF), an end-to-end agentic framework that closes this infrastructural gap. BTTF distills massive, unstructured simulation dumps into a normalized relational SQLite database and couples it with a collaborative multi-agent orchestration engine that translates natural-language verification queries into schema-aware SQL while correlating signal anomalies with versioned RTL repositories. Across a 150-query benchmark, BTTF attains 95.33% execution accuracy, charting a practical path toward autonomous EDA verification.
Comments40th NeurIPS'26 Workshop AI for Chip Design
