A study on curvilinear design and its impact on mask and wafer manufacturing

Optical proximity correction (OPC) is used to modulate the mask design to compensate the image error from the mask to the wafer. As the design feature size becomes smaller as the technology node progresses, OPC has become a key technology to lesson the process variability concern in semiconductor manufacturing. Recently, with a full adoption of EUV lithography and birth of high-NA EUV lithography with its added complexity compared to the DUV lithography, current computational lithography has hit the data wall that requires a great number of computation resources such as CPU cores, and this causes longer Turn-Around-Time (TAT) issue in manufacturing. Also, inverse lithography has been proven to be an efficient way to improve the process window, however, lack of computational resource has made it difficult to be adopted into the manufacturing. Considering above-mentioned challenges, machine learning has become a potential solution to dramatically reduce the computation time and resource.  In this work, the candidate will deep dive computational lithography engineering including resolution enhancement techniques (RET) and try to combine machine learning approach to seek feasible adoption of emerging computational lithography techniques.