用AI把自然语言设计需求自动转为光子芯片布局文件
AI Agents for Photonic Integrated Circuit Design Automation
- 多智能体系统将文字描述转化为芯片版图
- 15个以下元件的设计成功率达57%以上
- 适合光子芯片设计初学者和自动化研发团队
我们提出PhIDO，一种多智能体框架，可将自然语言的光子集成电路（PIC）设计请求转换为版图掩模文件。在包含102个设计描述的测试集上，对比了7种推理型大模型的表现，涵盖从单器件到112组件的复杂度。单器件设计的成功率最高达91%。对于组件数≤15的设计，o1、Gemini-2.5-pro和Claude Opus 4在端到端pass@5成功率上表现最优，约57%，其中Gemini-2.5-pro所需输出标记数最少且成本最低。未来迈向全自动光子芯片开发的关键包括：标准化知识表示、扩展数据集、增强验证能力及机器人自动化。
原文摘要 · Abstract (English)
We present Photonics Intelligent Design and Optimization (PhIDO), a multi-agent framework that converts natural-language photonic integrated circuit (PIC) design requests into layout mask files. We compare 7 reasoning large language models for PhIDO using a testbench of 102 design descriptions that ranged from single devices to 112-component PICs. The success rate for single-device designs was up to 91%. For design queries with less than or equal to 15 components, o1, Gemini-2.5-pro, and Claude Opus 4 achieved the highest end-to-end pass@5 success rates of approximately 57%, with Gemini-2.5-pro requiring the fewest output tokens and lowest cost. The next steps toward autonomous PIC development include standardized knowledge representations, expanded datasets, extended verification, and robotic automation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。