Extreme Ultraviolet Lithography
Extreme ultraviolet lithography (EUVL) is a critical technology for creating ever-smaller features on semiconductor chips, aiming to overcome the resolution limits of previous methods. Current research focuses on improving EUVL's accuracy and efficiency through machine learning models, including neural networks for lithography simulation and defect detection, and evolutionary algorithms for optimizing mask design and multiple patterning layout decomposition. These advancements address challenges like stochastic defects, process variations, and computational cost, ultimately impacting the fabrication of advanced integrated circuits and enabling continued progress in Moore's Law.
Papers
Characterization and Optimization of Integrated Silicon-Photonic Neural Networks under Fabrication-Process Variations
Asif Mirza, Amin Shafiee, Sanmitra Banerjee, Krishnendu Chakrabarty, Sudeep Pasricha, Mahdi Nikdast
A Thin Format Vision-Based Tactile Sensor with A Micro Lens Array (MLA)
Xia Chen, Guanlan Zhang, Michael Yu Wang, Hongyu Yu