Exploring Rule-Free Layout Decomposition via Deep Reinforcement Learning

Bentian Jiang*, Xinshi Zang, Martin D.F. Wong, Evangeline F.Y. Young

*Corresponding author for this work

Research output: Contribution to journalJournal articlepeer-review

1 Citation (Scopus)


Multiple patterning lithography decomposition (MPLD) and mask optimization enable the ever-shrinking device feature sizes far below the lithography system limit. Conventional MPLD is solved by mathematical programming or graph-based approaches, where a set of predetermined rules is indispensable to identify the conflicts to be resolved. In this article, we explore rule-free layout decomposition following a simple but sweet principle, let the mask optimizer 'teach' the layout decomposer how to generate suitable decompositions. Our flow includes a reinforcement-learning-based layout decomposer and a deep-learning-based mask optimizer. Without any handcrafted rules, our framework can perform competitively and even surpass the state-of-the-art rule-based methods with notable (7×∼ 63×) turn-around-time speedup.

Original languageEnglish
Pages (from-to)3067-3077
Number of pages11
JournalIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Issue number9
Early online date29 Dec 2022
Publication statusPublished - Sept 2023

Scopus Subject Areas

  • Software
  • Computer Graphics and Computer-Aided Design
  • Electrical and Electronic Engineering

User-Defined Keywords

  • Design for manufacturability
  • double patterning
  • inverse lithography technique (ILT)
  • reinforcement learning (RL)


Dive into the research topics of 'Exploring Rule-Free Layout Decomposition via Deep Reinforcement Learning'. Together they form a unique fingerprint.

Cite this