Assessment of Reinforcement Learning for Macro Placement

被引：13

作者：

Cheng, Chung-Kuan ^{[1
]}

Kahng, Andrew B. ^{[1
]}

Kundu, Sayak ^{[1
]}

Wang, Yucheng ^{[1
]}

Wang, Zhiang ^{[1
]}

机构：

[1] Univ Calif San Diego, La Jolla, CA 92093 USA

来源：

PROCEEDINGS OF THE 2023 INTERNATIONAL SYMPOSIUM ON PHYSICAL DESIGN, ISPD 2023 | 2023年

关键词：

Macro placement; Reinforcement learning; Modern benchmarks;

D O I：

10.1145/3569052.3578926

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

We provide open, transparent implementation and assessment of Google Brain's deep reinforcement learning approach to macro placement [9] and its Circuit Training (CT) implementation in Git-Hub [23]. We implement in open-source key "blackbox" elements of CT, and clarify discrepancies between CT and [9]. New testcases on open enablements are developed and released. We assess CT alongside multiple alternative macro placers, with all evaluation flows and related scripts public in GitHub. Our experiments also encompass academic mixed-size placement benchmarks, as well as ablation and stability studies. We comment on the impact of [9] and CT, as well as directions for future research.

引用

页码：158 / 166

页数：9

共 50 条

[1] Delving into Macro Placement with Reinforcement Learning
Jiang, Zixuan
Songhori, Ebrahim
Wang, Shen
Goldie, Anna
Mirhoseini, Azalia
Jiang, Joe
Lee, Young-Joon
Pan, David Z.
2021 ACM/IEEE 3RD WORKSHOP ON MACHINE LEARNING FOR CAD (MLCAD), 2021,
[2] Reevaluating Google's Reinforcement Learning for IC Macro Placement
Markov, Igor l.
COMMUNICATIONS OF THE ACM, 2024, 67 (11) : 60 - 71
[3] Hierarchical reinforcement learning for chip-macro placement in integrated circuit
Tan, Zhentao
Mu, Yadong
PATTERN RECOGNITION LETTERS, 2024, 179 : 108 - 114
[4] Non-overlapping placement of macro cells based on reinforcement learning in chip design
Yu, Tao
Gao, Peng
Wang, Fei
Yuan, Ru-Yue
INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS, 2024,
[5] Learning macro-actions in reinforcement learning
Randlov, J
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 11, 1999, 11 : 1045 - 1051
[6] Device Placement Optimization with Reinforcement Learning
Mirhoseini, Azalia
Pham, Hieu
Le, Quoc, V
Steiner, Benoit
Larsen, Rasmus
Zhou, Yuefeng
Kumar, Naveen
Norouzi, Mohammad
Bengio, Samy
Dean, Jeff
INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 70, 2017, 70
[7] Reusability and Transferability of Macro Actions for Reinforcement Learning
Chang Y.-H.
Chang K.-Y.
Kuo H.
Lee C.-Y.
ACM Transactions on Evolutionary Learning and Optimization, 2022, 2 (01):
[8] Hierarchical reinforcement learning based on macro actions
Hao Jiang
Gongju Wang
Shengze Li
Jieyuan Zhang
Long Yan
Xinhai Xu
Complex & Intelligent Systems, 2025, 11 (6)
[9] FPGA Placement Optimization with Deep Reinforcement Learning
Zhang, Junpeng
Deng, Fang
Yang, Xudong
2021 2ND INTERNATIONAL CONFERENCE ON COMPUTER ENGINEERING AND INTELLIGENT CONTROL (ICCEIC 2021), 2021, : 73 - 76
[10] MaMiC: Macro and Micro Curriculum for Robotic Reinforcement Learning
Tomar, Manan
Sathuluri, Akhil
Ravindran, Balaraman
AAMAS '19: PROCEEDINGS OF THE 18TH INTERNATIONAL CONFERENCE ON AUTONOMOUS AGENTS AND MULTIAGENT SYSTEMS, 2019, : 2226 - 2228

← 1 2 3 4 5 →