ML for EDA: Learning Algorithms for Analog and RF Physical Design

In the ever-evolving landscape of Electronic Design Automation (EDA), the integration of Artificial Intelligence (AI) and
Machine Learning (ML) algorithms with traditional heuristic optimization algorithms has emerged as a transformative force in
automated circuit design. This presentation delves into the dynamic intersection of AI/ML and EDA, exploring state-of-the-art
techniques shaping the analog and RF physical design space. Machine learning, including deep learning, has the potential to
significantly improve the accuracy, speed, efficiency, and reliability of EDA tasks such as circuit modeling, simulation, layout
design, and optimization. Delving into such cutting-edge advancements, I will describe current AI/ML research performed by the
Drexel ICE Lab that extends beyond traditional analog and RF design paradigms, with the goal of enabling designers to navigate
complexities with unparalleled efficiency and accuracy. Specifically, a focus on state-of-the-art learning and optimization
techniques for the modeling and design of analog and RF ICs will be presented and described. Practical considerations,
challenges, and opportunities of ML algorithms for analog and RF circuit design will be discussed, with a focus on the use of
such algorithms for prediction and optimization tasks within the EDA design flow.