AI Agents for Photonic Integrated Circuit Design Automation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913998113341440 |
|---|---|
| author | Sharma, Ankita Fu, YuQi Ansari, Vahid Iyer, Rishabh Kuang, Fiona Mistry, Kashish Aishy, Raisa Islam Ahmad, Sara Matres, Joaquin Englund, Dirk R. Poon, Joyce K. S. |
| author_facet | Sharma, Ankita Fu, YuQi Ansari, Vahid Iyer, Rishabh Kuang, Fiona Mistry, Kashish Aishy, Raisa Islam Ahmad, Sara Matres, Joaquin Englund, Dirk R. Poon, Joyce K. S. |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_14123 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | AI Agents for Photonic Integrated Circuit Design Automation Sharma, Ankita Fu, YuQi Ansari, Vahid Iyer, Rishabh Kuang, Fiona Mistry, Kashish Aishy, Raisa Islam Ahmad, Sara Matres, Joaquin Englund, Dirk R. Poon, Joyce K. S. Hardware Architecture Artificial Intelligence Applied Physics Optics 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. |
| title | AI Agents for Photonic Integrated Circuit Design Automation |
| topic | Hardware Architecture Artificial Intelligence Applied Physics Optics |
| url | https://arxiv.org/abs/2508.14123 |