Prof. Hua Wang

Speaker: Prof. Hua Wang (ETH, Zurich, Swiss)

Title: AI-Assisted for Radio-Frequency Integrated Circuits at ETH Zurich

Abstract: In recent years, as AI for Science is gaining increasing momentum, there are rapidly growing interest in applying AI tools to facilitate RFIC designs or explore new design spaces. In this seminar talk, I will present a few on-going AI4RF projects at my IDEAS group at ETH Zurich. We will start with our project of AI-assisted template-seeded pixelated design for multi-metal-layer EM structures, which explores a new approach for efficient exploration and exploitation of the design space. Next, we will introduce our transfer learning assisted fast design migration, which potentially enables efficient design migrations over technology nodes and operating frequencies. Finally, we will present out project on top-metal-only RFIC redesigns for fast specs-to-silicon iteration enabled by AI-assisted inverse design. Finally, we will discuss challenges and future directions.