• Technical Name
  • From IC Layout to Die Photo: Deep Learning-Based IC Fabrication SimulationMask Correction
  • Operator
  • National Tsing Hua University
  • Booth
  • Online display only
  • Contact
  • 劉卉宥
  • Email
  • jsyoaliu@gmail.com
Technical Description This technology consists of two convolutional neural networks (CNNs): LithoNetOPCNet. LithoNet mimics the fabrication procedure, involving lithographyetching, to predict the shape of a fabricated IC circuitry given an input layout. OPCNet, in cooperation with a pre-trained LithoNet, mimics the optical proximity correction procedure used to correct the layout designgenerate a modif
Scientific Breakthrough This technology is the first to use a deep neural network to accurately model the effect of lithographyetching processes on the shape deformation of a fabricated IC circuitry. It can also generates a correct mask pattern so that the fabricated IC circuitry will have a shape close to a desired pattern. The technolgy can lead to a paradigm shift of IC design for manufacturability.
Industrial Applicability This technology can be used by semiconductor manufacturers for modeling IC fabrication processesIC design CAD tools for design for Manufacturability.
本網站使用您的Cookie於優化網站。繼續瀏覽網站即表示您同意本公司隱私權政策,您可至隱私權政策了解詳細資訊。