The department of AI for EDA is responsible for development and integration of AI and ML algorithms with applications in Electronic Design Automation. Hardware design is suffering from a talent gap, where the number of new hardware engineers cannot keep up with the ever-increasing demand for faster and more powerful chips. As a result, hardware engineers are under pressure to design more complex and more powerful hardware in shorter periods of time. The hardware design field is ripe for AI disruption to assist hardware engineers with improved productivity and faster design cycles. AI for EDA department at IBM Research responsibilities cover the entire spectrum of design and takes advantage of technologies such as Deep Learning, Graph Neural Networks, Transformers, Large Language Models and Reinforcement Learning. Current projects include application of Large Language Models for coding and assisting chatbots,distributed optimization for digital design flow and analog/mixed signal circuit parameters, graph neural networks for prediction of design behavior and reinforcement learning for macro-placement. These technologies can assist hardware designers in many steps of their design workflow to improve their productivity.
The members of the AI for EDA team are located at IBM Almaden Research Center.They have experience and expertise in a broad range of topics including Computer Vision, Natural Language Processing, Distributed Machine Learning and Generative AI, in addition to Electronic Design Automation.
