PhD - Leuven | More than two weeks ago
Optical proximity correction (OPC) is used to modulate the mask design to compensate the image error from the mask to the wafer, and design for manufacturability (DFM) to further assist OPC by modifying the design into fully manufacturable form. As the design feature size becomes smaller as the technology node progresses, OPC and DFM have become critical technologies in semiconductor manufacturing. However, recently, with a full adoption of EUV lithography into manufacturing and the birth of high-NA EUV lithography, current computational lithography is hitting a wall in processing the excessive amount of data (i.e. big data) in the model and algorithm as it requires a great deal of computation resources such as CPU cores, and this causes longer than expected turn-around-time (TAT) that puts challenge in high volume manufacturing of semiconductor devices. Also, inverse lithography technique (ILT) has been proven to be a promising method to further improve the process window and manufacturability as an extension of conventional OPC, however, big data handling issue during the computation has made it difficult to be fully acceptable yet. Considering above-mentioned challenges, machine learning has become potential and promising solution which is expected to reduce the complexity if used in combination with above-mentioned novel technologies.
In this work, the candidate will deep dive machine learning to seek feasible adoption of such emerging computational lithography techniques.
Required background: Engineering technology
Type of work: 100% modeling/simulation
Supervisor: Stefan De Gendt
Co-supervisor: Ryan Ryoung han Kim
Daily advisor: Apoorva Oak
The reference code for this position is 2021-050. Mention this reference code on your application form.