Mitigating Distribution Shift for Congestion Optimization in Global Placement

Su Zheng; Lancheng Zou; Siting Liu; Yibo Lin; Bei Yu; Martin Wong

doi:10.1109/DAC56929.2023.10247660

Mitigating Distribution Shift for Congestion Optimization in Global Placement

Su Zheng, Lancheng Zou, Siting Liu, Yibo Lin, Bei Yu, Martin Wong

Office of the Provost

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

3 Citations (Scopus)

Abstract

The placement and routing (PnR) flow plays a critical role in physical design. Poor routing congestion is a possible problem causing severe routing detours, which can lead to deteriorated timing performance or even routing failure. Deep-learning-based congestion prediction model is designed to guide the global placement process in previous work. However, the distribution shift problem in this method limits its performance. In this paper, we mitigate the distribution shift problem with a look-ahead mechanism inspired by optical flow prediction and an invariant feature space learning technique. With the proposed method, we can achieve better congestion prediction performance and less-congested placement results.

Original language	English
Title of host publication	60th ACM/IEEE Design Automation Conference - Proceedings 2023
Publisher	IEEE
Pages	1-6
Number of pages	6
ISBN (Electronic)	9798350323481
ISBN (Print)	9798350323498
DOIs	https://doi.org/10.1109/DAC56929.2023.10247660
Publication status	Published - 13 Jul 2023
Event	60th ACM/IEEE Design Automation Conference, DAC 2023 - Moscone West, San Francisco, United States Duration: 9 Jul 2023 → 13 Jul 2023 https://www.dac.com/ https://60dac.conference-program.com/ https://ieeexplore.ieee.org/xpl/conhome/10247654/proceeding

Publication series

Name	ACM/IEEE Design Automation Conference - Proceedings
Volume	2023-July
ISSN (Print)	0738-100X

Conference

Conference	60th ACM/IEEE Design Automation Conference, DAC 2023
Country/Territory	United States
City	San Francisco
Period	9/07/23 → 13/07/23
Internet address	https://www.dac.com/ https://60dac.conference-program.com/ https://ieeexplore.ieee.org/xpl/conhome/10247654/proceeding

Scopus Subject Areas

Electrical and Electronic Engineering
Control and Systems Engineering
Computer Science Applications
Modelling and Simulation

User-Defined Keywords

Benchmark testing
Design automation
Optical flow
Optimization methods
Predictive models
Routing
Timing

Access to Document

10.1109/DAC56929.2023.10247660

Cite this

@inproceedings{cfcc778d572f4f1484a23c45fd1ef535,

title = "Mitigating Distribution Shift for Congestion Optimization in Global Placement",

abstract = "The placement and routing (PnR) flow plays a critical role in physical design. Poor routing congestion is a possible problem causing severe routing detours, which can lead to deteriorated timing performance or even routing failure. Deep-learning-based congestion prediction model is designed to guide the global placement process in previous work. However, the distribution shift problem in this method limits its performance. In this paper, we mitigate the distribution shift problem with a look-ahead mechanism inspired by optical flow prediction and an invariant feature space learning technique. With the proposed method, we can achieve better congestion prediction performance and less-congested placement results.",

keywords = "Benchmark testing, Design automation, Optical flow, Optimization methods, Predictive models, Routing, Timing",

author = "Su Zheng and Lancheng Zou and Siting Liu and Yibo Lin and Bei Yu and Martin Wong",

note = "Funding Information: This work is supported by The Research Grants Council of Hong Kong SAR (No. CUHK14209420). Publisher Copyright: {\textcopyright} 2023 IEEE.; 60th ACM/IEEE Design Automation Conference, DAC 2023 ; Conference date: 09-07-2023 Through 13-07-2023",

year = "2023",

month = jul,

day = "13",

doi = "10.1109/DAC56929.2023.10247660",

language = "English",

isbn = "9798350323498",

series = "ACM/IEEE Design Automation Conference - Proceedings",

publisher = "IEEE",

pages = "1--6",

booktitle = "60th ACM/IEEE Design Automation Conference - Proceedings 2023",

address = "United States",

url = "https://www.dac.com/, https://60dac.conference-program.com/, https://ieeexplore.ieee.org/xpl/conhome/10247654/proceeding",

}

Zheng, S, Zou, L, Liu, S, Lin, Y, Yu, B & Wong, M 2023, Mitigating Distribution Shift for Congestion Optimization in Global Placement. in 60th ACM/IEEE Design Automation Conference - Proceedings 2023. ACM/IEEE Design Automation Conference - Proceedings, vol. 2023-July, IEEE, pp. 1-6, 60th ACM/IEEE Design Automation Conference, DAC 2023, San Francisco, California, United States, 9/07/23. https://doi.org/10.1109/DAC56929.2023.10247660

Mitigating Distribution Shift for Congestion Optimization in Global Placement. / Zheng, Su; Zou, Lancheng; Liu, Siting et al.
60th ACM/IEEE Design Automation Conference - Proceedings 2023. IEEE, 2023. p. 1-6 (ACM/IEEE Design Automation Conference - Proceedings; Vol. 2023-July).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Mitigating Distribution Shift for Congestion Optimization in Global Placement

AU - Zheng, Su

AU - Zou, Lancheng

AU - Liu, Siting

AU - Lin, Yibo

AU - Yu, Bei

AU - Wong, Martin

PY - 2023/7/13

Y1 - 2023/7/13

N2 - The placement and routing (PnR) flow plays a critical role in physical design. Poor routing congestion is a possible problem causing severe routing detours, which can lead to deteriorated timing performance or even routing failure. Deep-learning-based congestion prediction model is designed to guide the global placement process in previous work. However, the distribution shift problem in this method limits its performance. In this paper, we mitigate the distribution shift problem with a look-ahead mechanism inspired by optical flow prediction and an invariant feature space learning technique. With the proposed method, we can achieve better congestion prediction performance and less-congested placement results.

AB - The placement and routing (PnR) flow plays a critical role in physical design. Poor routing congestion is a possible problem causing severe routing detours, which can lead to deteriorated timing performance or even routing failure. Deep-learning-based congestion prediction model is designed to guide the global placement process in previous work. However, the distribution shift problem in this method limits its performance. In this paper, we mitigate the distribution shift problem with a look-ahead mechanism inspired by optical flow prediction and an invariant feature space learning technique. With the proposed method, we can achieve better congestion prediction performance and less-congested placement results.

KW - Benchmark testing

KW - Design automation

KW - Optical flow

KW - Optimization methods

KW - Predictive models

KW - Routing

KW - Timing

UR - https://ieeexplore.ieee.org/document/10247660/

UR - http://www.scopus.com/inward/record.url?scp=85173079089&partnerID=8YFLogxK

U2 - 10.1109/DAC56929.2023.10247660

DO - 10.1109/DAC56929.2023.10247660

M3 - Conference proceeding

SN - 9798350323498

T3 - ACM/IEEE Design Automation Conference - Proceedings

SP - 1

EP - 6

BT - 60th ACM/IEEE Design Automation Conference - Proceedings 2023

PB - IEEE

T2 - 60th ACM/IEEE Design Automation Conference, DAC 2023

Y2 - 9 July 2023 through 13 July 2023

ER -

Mitigating Distribution Shift for Congestion Optimization in Global Placement

Abstract

Publication series

Conference

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this