Prediction of Cross-Sectional Images and Proposing Processes with Neural Network

We developed a novel system that quickly predicts cross-sectional images from experimental conditions to minimize cross-sectional processing. This has become a bottleneck in observing pattern shapes of materials used in substrate manufacturing. This system can predict cross-sectional images from experimental conditions and propose process conditions to form the desired pattern shape. We consider that using this system will improve the development speed of overall semiconductor packages.