Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2407.03792 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909242390216704 |
|---|---|
| author | Manchanda, Sahil Kianfar, Dana Peschl, Markus Lepert, Romain Defferrard, Michaël |
| author_facet | Manchanda, Sahil Kianfar, Dana Peschl, Markus Lepert, Romain Defferrard, Michaël |
| contents | A core objective of physical design is to minimize wirelength (WL) when placing chip components on a canvas. Computing the minimal WL of a placement requires finding rectilinear Steiner minimum trees (RSMTs), an NP-hard problem. We propose NeuroSteiner, a neural model that distills GeoSteiner, an optimal RSMT solver, to navigate the cost--accuracy frontier of WL estimation. NeuroSteiner is trained on synthesized nets labeled by GeoSteiner, alleviating the need to train on real chip designs. Moreover, NeuroSteiner's differentiability allows to place by minimizing WL through gradient descent. On ISPD 2005 and 2019, NeuroSteiner can obtain 0.3% WL error while being 60% faster than GeoSteiner, or 0.2% and 30%. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_03792 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | NeuroSteiner: A Graph Transformer for Wirelength Estimation Manchanda, Sahil Kianfar, Dana Peschl, Markus Lepert, Romain Defferrard, Michaël Machine Learning A core objective of physical design is to minimize wirelength (WL) when placing chip components on a canvas. Computing the minimal WL of a placement requires finding rectilinear Steiner minimum trees (RSMTs), an NP-hard problem. We propose NeuroSteiner, a neural model that distills GeoSteiner, an optimal RSMT solver, to navigate the cost--accuracy frontier of WL estimation. NeuroSteiner is trained on synthesized nets labeled by GeoSteiner, alleviating the need to train on real chip designs. Moreover, NeuroSteiner's differentiability allows to place by minimizing WL through gradient descent. On ISPD 2005 and 2019, NeuroSteiner can obtain 0.3% WL error while being 60% faster than GeoSteiner, or 0.2% and 30%. |
| title | NeuroSteiner: A Graph Transformer for Wirelength Estimation |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2407.03792 |