TY - GEN
T1 - Physics-Informed Optical Kernel Regression Using Complex-valued Neural Fields
AU - Chen, Guojin
AU - Pei, Zehua
AU - Yang, Haoyu
AU - Ma, Yuzhe
AU - Yu, Bei
AU - Wong, Martin
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Lithography is fundamental to integrated circuit fabrication, necessitating large computation overhead. The advancement of machine learning (ML)-based lithography models alleviates the trade-offs between manufacturing process expense and capability. However, all previous methods regard the lithography system as an image-to-image black box mapping, utilizing network parameters to learn by rote mappings from massive mask-to-aerial or mask-to-resist image pairs, resulting in poor generalization capability. In this paper, we propose a new ML-based paradigm disassembling the rigorous lithographic model into non-parametric mask operations and learned optical kernels containing determinant source, pupil, and lithography information. By optimizing complex-valued neural fields to perform optical kernel regression from coordinates, our method can accurately restore lithography system using a small-scale training dataset with fewer parameters, demonstrating superior generalization capability as well. Experiments show that our framework can use 31% of parameters while achieving 69× smaller mean squared error with 1.3× higher throughput than the state-of-the-art.
AB - Lithography is fundamental to integrated circuit fabrication, necessitating large computation overhead. The advancement of machine learning (ML)-based lithography models alleviates the trade-offs between manufacturing process expense and capability. However, all previous methods regard the lithography system as an image-to-image black box mapping, utilizing network parameters to learn by rote mappings from massive mask-to-aerial or mask-to-resist image pairs, resulting in poor generalization capability. In this paper, we propose a new ML-based paradigm disassembling the rigorous lithographic model into non-parametric mask operations and learned optical kernels containing determinant source, pupil, and lithography information. By optimizing complex-valued neural fields to perform optical kernel regression from coordinates, our method can accurately restore lithography system using a small-scale training dataset with fewer parameters, demonstrating superior generalization capability as well. Experiments show that our framework can use 31% of parameters while achieving 69× smaller mean squared error with 1.3× higher throughput than the state-of-the-art.
UR - http://www.scopus.com/inward/record.url?scp=85173072779&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85173072779&partnerID=8YFLogxK
U2 - 10.1109/DAC56929.2023.10247680
DO - 10.1109/DAC56929.2023.10247680
M3 - Conference contribution
AN - SCOPUS:85173072779
T3 - Proceedings - Design Automation Conference
BT - 2023 60th ACM/IEEE Design Automation Conference, DAC 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 60th ACM/IEEE Design Automation Conference, DAC 2023
Y2 - 9 July 2023 through 13 July 2023
ER -