MaskOpt: A Large-Scale Mask Optimization Dataset to Advance AI in Integrated Circuit Manufacturing
Published:Dec 25, 2025 05:00
•1 min read
•ArXiv ML
Analysis
This paper introduces MaskOpt, a new large-scale dataset designed to improve the application of deep learning in integrated circuit (IC) mask optimization. The dataset addresses limitations in existing datasets by using real IC designs at the 45nm node, incorporating standard-cell hierarchy, and considering surrounding contexts. The authors emphasize the importance of these factors for practical mask optimization. By providing a benchmark for cell- and context-aware mask optimization, MaskOpt aims to facilitate the development of more effective deep learning models. The paper includes an evaluation of state-of-the-art models and analysis of context size and input ablation, highlighting the dataset's utility and potential impact on the field. The focus on real-world data and practical considerations makes this a valuable contribution.
Key Takeaways
- •Introduces MaskOpt, a large-scale dataset for IC mask optimization.
- •Uses real IC designs at the 45nm node.
- •Focuses on cell- and context-aware mask optimization.
Reference
“To advance deep learning for cell- and context-aware mask optimization, we present MaskOpt, a large-scale benchmark dataset constructed from real IC designs at the 45$\mathrm{nm}$ node.”