Deep Learning-Based Early-Stage IR-Drop Estimation via CNN Surrogate Modeling
Ritesh Bhadana · 2026
Read Deep Learning-Based Early-Stage IR-Drop Estimation via CNN Surrogate Modeling by Ritesh Bhadana (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-2601-22707.txt.
Description
Ritesh Bhadana predicts voltage droop on a chip early, with a convolutional stand-in for costly physics tools. IR-drop, as the droop is called, can slow a chip, wear it out, or break it. His 2026 preprint notes that sign-off tools are exact but slow, and need a near-final layout.
A CNN surrogate can flag droop while the design is still young and cheap to change. Early flags save respins, which cost firms dearly. The work fits the push for AI in chip design. It trades a little exactness for much speed. That trade, early, is wise.
Catalog record
| CREATOR | Ritesh Bhadana |
|---|---|
| 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/2601-22707