AI-based EDA Tools For 3D-IC Physical Design (AI-ICPD) (ART/374CP)

The global demand for 3D-IC is growing rapidly, and related applications range from edge computing, the Internet of things, data centers, AI computing, and automotive. The size of the global 3D-IC market is expected to reach $52 billion in 2030, with a CAGR of about 20% from 2021 to 2030. To take full advantage of 3D-IC, sophisticated design techniques, and new EDA tools are required. EDA plays a crucial role in chip design from the initial concept to the final signature. The global EDA market is expected to reach $22 billion in 2028, with a CAGR of about 7% from 2021 to 2028. EDA tools, especially 3D-IC EDA tools, have a vast market potential. AI can empower EDA to tackle complexity, cost, and time-to-market challenges encountered in 3D-IC design. This project will develop an AI-based 3D-IC physical design platform that specifically includes the development of floor plan, placement and routing algorithms, and the integrated EDA software with customizable API. To achieve this, we will use the most advanced artificial intelligence technology to develop floor plan, placement, and routing algorithms with TSV (Through Silicon Via) awareness to optimize performance, power consumption, and area in chip design, leverage heterogeneous computing resources, and develop EDA software with customizable API that can easily be integrated with other commercial EDA tools. The developed solutions will help domestic EDA companies fill the gap in the EDA tools for 3D-IC physical design.