Wang, M. orcid.org/0009-0005-6222-1778, Cheng, Y. orcid.org/0000-0003-2477-314X, Lin, Y. orcid.org/0009-0005-2158-3027 et al. (4 more authors) (2024) MAUnet: Multiscale attention U-Net for effective IR drop prediction. In: DAC '24: Proceedings of the 61st ACM/IEEE Design Automation Conference. DAC '24: 61st ACM/IEEE Design Automation Conference, 23-27 Jun 2024, San Francisco CA, USA. ACM , pp. 1-6. ISBN 9798400706011
Abstract
The efficient analysis of power grids is a crucial yet computationally challenging task in integrated circuit (IC) design, given the shrinking power supply voltage of ultra deep-submicron VLSI design. Different from the conventional modified nodal analysis technique, this paper introduces MAUnet, an innovative machine-learning model that redefines state-of-the-art full-chip static IR drop prediction. MAUnet ingeniously integrates multi-scale convolutional blocks, attention mechanisms, and U-Net architecture to optimize prediction accuracy. The multi-scale convolutional blocks significantly enhance feature extraction from image-based data, while the attention mechanism precisely identifies hotspot regions. The U-Net architecture, on the other hand, enables scalable image-to-image prediction applicable to circuits of any size. Uniquely, MAUnet also incorporates a pioneering fusion method that synergies both power grids and image-based data. Additionally, we introduce a low-rank approximation transfer learning technique to extend MAUnet's applicability to unseen test cases. Benchmark tests validate MAUnet's superior performance, achieving an average error of less than 6% relative to the average IR drop on three benchmarks. The performance enhancements offered by our proposed method are substantial, outperforming the current state-of-the-art method, IREDGe, by considerable margins of 29%, 65%, and 68% in three canonical benchmarks. Transfer learning is validated to enable model to achieve effective improvement on real circuit test cases. Compared to commercial tools, which often require hours to deliver results, the proposed method provides orders of magnitude speed-up with negligible error in practice.
Metadata
Item Type: | Proceedings Paper |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | © 2024 Copyright held by the owner/author(s).| ACM 2024. This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in DAC '24: Proceedings of the 61st ACM/IEEE Design Automation Conference, http://dx.doi.org/10.1145/3649329.3658465 |
Keywords: | IR drop analysis; U-net; attention mechanism; machine learning |
Dates: |
|
Institution: | The University of Sheffield |
Academic Units: | The University of Sheffield > Faculty of Science (Sheffield) > School of Mathematical and Physical Sciences |
Depositing User: | Symplectic Sheffield |
Date Deposited: | 19 Dec 2024 10:55 |
Last Modified: | 19 Dec 2024 11:04 |
Status: | Published |
Publisher: | ACM |
Refereed: | Yes |
Identification Number: | 10.1145/3649329.3658465 |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:221017 |