GNN-based Path-aware multi-view Circuit Learning for Technology Mapping
Wentao Jiang *, Jingxin Wang *, Zhang Hu, Zhengyuan Shi, Chengyu Ma, Qiang Xu, Weikang Qian, Zhufei Chu † · 2026
Read GNN-based Path-aware multi-view Circuit Learning for Technology Mapping by Wentao Jiang *, Jingxin Wang *, Zhang Hu, Zhengyuan Shi, Chengyu Ma, Qiang Xu, Weikang Qian, Zhufei Chu † (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-2601-14286.txt.
Description
Wentao Jiang, Jingxin Wang, and six co-authors work on chip design in this 2026 paper. Their stage is technology mapping, the step that turns a logic plan into real gates from a cell library. Old mapping tools lean on abstract delay models. Those models do not tie to a real target technology.
That gap leads to wide errors in delay estimates for the built circuit. Real post-mapping circuits show timing traits the abstract models miss. The team built a learning method based on graph neural networks. It is path-aware and uses multi-view circuit learning. The aim is mapping choices grounded in how real circuits time their signals.
Catalog record
| CREATOR | Wentao Jiang *, Jingxin Wang *, Zhang Hu, Zhengyuan Shi, Chengyu Ma, Qiang Xu, Weikang Qian, Zhufei Chu † |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-14286