An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search

Jinghao Ding; Wenxin Yu; Yuanrui Qi; Zhaoqi Fu; Mengshi Gong; I. Chyn Wey; Jinjia Zhou

doi:10.1145/3716368.3735152

An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search

Jinghao Ding, Wenxin Yu^*, Yuanrui Qi, Zhaoqi Fu, Mengshi Gong, I. Chyn Wey, Jinjia Zhou

^*Corresponding author for this work

Department of Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Macro placement plays a crucial role in physical design, directly influencing power, performance, and area. While existing approaches, such as optimization-based and learning-based methods, have made progress, they still face significant challenges in terms of exploration efficiency and optimization capability. In this paper, we introduce an enhanced reinforcement learning (RL)-based framework integrated with Monte Carlo Tree Search (MCTS). A dynamic node selection strategy is used to optimize frontier nodes in MCTS for macro placement, while a curiosity-driven exploration mechanism generates intrinsic rewards to enhance the efficiency of exploring diverse placement solutions. Additionally, prioritized experience replay focuses on key placement states, further improving optimization performance. Experimental results on the ISPD2005 benchmark demonstrate that our placement wirelength outperforms recent state-of-the-art methods.

Original language	English
Title of host publication	GLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025
Publisher	Association for Computing Machinery
Pages	764-769
Number of pages	6
ISBN (Electronic)	9798400714962
DOIs	https://doi.org/10.1145/3716368.3735152
State	Published - 29 06 2025
Event	35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025 - New Orleans, United States Duration: 30 06 2025 → 02 07 2025

Publication series

Name	Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI

Conference

Conference	35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025
Country/Territory	United States
City	New Orleans
Period	30/06/25 → 02/07/25

Bibliographical note

Publisher Copyright:
© 2025 Copyright held by the owner/author(s).

Keywords

Macro Placement
Monte Carlo Tree Search
Reinforcement Learning

Access to Document

10.1145/3716368.3735152

Cite this

Ding, J., Yu, W., Qi, Y., Fu, Z., Gong, M., Wey, I. C., & Zhou, J. (2025). An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search. In GLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025 (pp. 764-769). (Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI). Association for Computing Machinery. https://doi.org/10.1145/3716368.3735152

@inproceedings{8dc5b3373a2745b1b5d1a237232884c5,

title = "An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search",

abstract = "Macro placement plays a crucial role in physical design, directly influencing power, performance, and area. While existing approaches, such as optimization-based and learning-based methods, have made progress, they still face significant challenges in terms of exploration efficiency and optimization capability. In this paper, we introduce an enhanced reinforcement learning (RL)-based framework integrated with Monte Carlo Tree Search (MCTS). A dynamic node selection strategy is used to optimize frontier nodes in MCTS for macro placement, while a curiosity-driven exploration mechanism generates intrinsic rewards to enhance the efficiency of exploring diverse placement solutions. Additionally, prioritized experience replay focuses on key placement states, further improving optimization performance. Experimental results on the ISPD2005 benchmark demonstrate that our placement wirelength outperforms recent state-of-the-art methods.",

keywords = "Macro Placement, Monte Carlo Tree Search, Reinforcement Learning",

author = "Jinghao Ding and Wenxin Yu and Yuanrui Qi and Zhaoqi Fu and Mengshi Gong and Wey, \{I. Chyn\} and Jinjia Zhou",

note = "Publisher Copyright: {\textcopyright} 2025 Copyright held by the owner/author(s).; 35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025 ; Conference date: 30-06-2025 Through 02-07-2025",

year = "2025",

month = jun,

day = "29",

doi = "10.1145/3716368.3735152",

language = "英语",

series = "Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI",

publisher = "Association for Computing Machinery",

pages = "764--769",

booktitle = "GLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025",

address = "美国",

}

Ding, J, Yu, W, Qi, Y, Fu, Z, Gong, M, Wey, IC & Zhou, J 2025, An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search. in GLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025. Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI, Association for Computing Machinery, pp. 764-769, 35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025, New Orleans, United States, 30/06/25. https://doi.org/10.1145/3716368.3735152

An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search. / Ding, Jinghao; Yu, Wenxin; Qi, Yuanrui et al.
GLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025. Association for Computing Machinery, 2025. p. 764-769 (Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search

AU - Ding, Jinghao

AU - Yu, Wenxin

AU - Qi, Yuanrui

AU - Fu, Zhaoqi

AU - Gong, Mengshi

AU - Wey, I. Chyn

AU - Zhou, Jinjia

PY - 2025/6/29

Y1 - 2025/6/29

N2 - Macro placement plays a crucial role in physical design, directly influencing power, performance, and area. While existing approaches, such as optimization-based and learning-based methods, have made progress, they still face significant challenges in terms of exploration efficiency and optimization capability. In this paper, we introduce an enhanced reinforcement learning (RL)-based framework integrated with Monte Carlo Tree Search (MCTS). A dynamic node selection strategy is used to optimize frontier nodes in MCTS for macro placement, while a curiosity-driven exploration mechanism generates intrinsic rewards to enhance the efficiency of exploring diverse placement solutions. Additionally, prioritized experience replay focuses on key placement states, further improving optimization performance. Experimental results on the ISPD2005 benchmark demonstrate that our placement wirelength outperforms recent state-of-the-art methods.

AB - Macro placement plays a crucial role in physical design, directly influencing power, performance, and area. While existing approaches, such as optimization-based and learning-based methods, have made progress, they still face significant challenges in terms of exploration efficiency and optimization capability. In this paper, we introduce an enhanced reinforcement learning (RL)-based framework integrated with Monte Carlo Tree Search (MCTS). A dynamic node selection strategy is used to optimize frontier nodes in MCTS for macro placement, while a curiosity-driven exploration mechanism generates intrinsic rewards to enhance the efficiency of exploring diverse placement solutions. Additionally, prioritized experience replay focuses on key placement states, further improving optimization performance. Experimental results on the ISPD2005 benchmark demonstrate that our placement wirelength outperforms recent state-of-the-art methods.

KW - Macro Placement

KW - Monte Carlo Tree Search

KW - Reinforcement Learning

UR - https://www.scopus.com/pages/publications/105017740111

U2 - 10.1145/3716368.3735152

DO - 10.1145/3716368.3735152

M3 - 会议稿件

AN - SCOPUS:105017740111

T3 - Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI

SP - 764

EP - 769

BT - GLSVLSI 2025 - Proceedings of the Great Lakes Symposium on VLSI 2025

PB - Association for Computing Machinery

T2 - 35th Edition of the Great Lakes Symposium on VLSI 2025, GLSVLSI 2025

Y2 - 30 June 2025 through 2 July 2025

ER -

An Effective Macro Placement Framework with Reinforcement Learning and Monte Carlo Tree Search

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this