TOPCELL: Topology Optimization of Standard Cell via LLMs Conference: ; July 26–29, 2026; Long Beach, CA Conference: 63rd ACM/IEEE Design Automation Conference; July 26–29, 2026; Long Beach, CA, USA 63rd ACM/IEEE Design Automation Conference (DAC ’26), July 26–29, 2026, Long Beach, CA, USA DOI: 10.1145/3770743.3804365 ISBN: 979-8-4007-2254-7/2026/07
Zhan Song, Yu-Tung Liu, Chen Chen, Guoheng Sun, Jiaqi Yin, Chia-tung Ho, Ang Li, Haoxing Ren, Cunxi Yu · 2026
Read TOPCELL: Topology Optimization of Standard Cell via LLMs Conference: ; July 26–29, 2026; Long Beach, CA Conference: 63rd ACM/IEEE Design Automation Conference; July 26–29, 2026; Long Beach, CA, USA 63rd ACM/IEEE Design Automation Conference (DAC ’26), July 26–29, 2026, Long Beach, CA, USA DOI: 10.1145/3770743.3804365 ISBN: 979-8-4007-2254-7/2026/07 by Zhan Song, Yu-Tung Liu, Chen Chen, Guoheng Sun, Jiaqi Yin, Chia-tung Ho, Ang Li, Haoxing Ren, Cunxi Yu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-14237.txt.
Description
Song and eight colleagues turn transistor topology search into a generative task for large language models with TOPCELL. Fine-tuned with group relative policy optimization, the model aligns its topology choices with circuit logic and layout space. In an industrial flow aimed at a 2nm node, TOPCELL beats foundation models at finding routable, physically aware topologies. Plugged into a state-of-the-art flow for 7nm library generation, it generalizes with no examples and matches exhaustive solvers on layout quality while running about 86 times faster.
Catalog record
| CREATOR | Zhan Song, Yu-Tung Liu, Chen Chen, Guoheng Sun, Jiaqi Yin, Chia-tung Ho, Ang Li, Haoxing Ren, Cunxi Yu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34837 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.14237