Reduction of systematic defects with machine learning from design to fab

This paper presents machine learning and computational lithography methods to identify and eliminate yield limiting patterns in the design, improve the accuracy of mask generation with etch and resist modeling and OPC, and improve the productivity and accuracy of fab defect detection and diagnostics.