ML-Powered Semiconductor Design Optimization

The Virtual Semiconductor Fabrication Model is a machine learning system developed by the National Research Council Canada in collaboration with McGill University. This digital solution helps reduce the impact of fabrication imperfections on semiconductor component optical performance by predicting how design layouts will perform during manufacturing. Designers can conduct virtual trials on the model to verify that client expectations will be met and gather valuable data before sending designs to manufacturers. The system uses deep learning trained on internal research data and fabrication/imaging data from Applied Nanotools Inc. Primary users are Government of Canada employees. This system does not involve personal information.