Deep learning-based EDA tools for lithography simulation, photomask correction, and novel layout patterns detection
Summary The DNN models of this technology include a LithoNet, an OPCNet, and a layout novelty detection network. LithoNet is a learning-based pre-simulation model for layout-to-SEM contour prediction, and OPCnet is a dual network of LithoNet for photomask optimization. Integrated with a well-trained LithoNet, our layout novelty detection network, consisting of a self-attention guided LithoNet and an autoencoder, can check if there are layout patterns easily resulting in local distortions in contours of metal lines based on multi-modal (global-local) feature fusion.

This is the first set of image-based layout-to-SEM prediction, photomask optimization, and layout novelty detection methods. Given different training ADI/AEI images, this technology can learn the knowledge of photolithograph and etching effects. It can thus help semiconductor manufacturers to predict the shape of metal layers of IC products and detect layout novelties. Also, this learning-based technology can be easily extended by collecting suitable layout-SEM image pairs of different fabrication configurations. Hence, it sets a new milestone and breakthrough for EDA tool development.

Compared with existing commercial rule-based software designed for analyzing ADI (after-development-image), the proposed technology comprising three learning-based methods can easily learn etching effects by training them on AEI (after-etching-image) image datasets. Therefore, this technology can also be used to predict layout-to-SEM contour deformations, optimized photo-masks, and layout novelty/weakness with respect to various fabrication parameters. Through this way, this technology can assist semiconductor manufacturers to save testing time and fabrication costs.
Technical Film
Keyword Optoelectronic Semiconductor Technology Interdisciplinary integration
Provide the latest information of AI research centers and applied industries