Inserting AI and machine learning into chips adds a whole new dimension of complexity, and creates a variety of potential problems, including deadlocks, loss of performance, and difficulty in achieving closure on many fronts. Gajinder Panesar, fellow at Siemens EDA, talks with Semiconductor Engineering about what’s changed and how to optimize these new devices and systems by monitoring them from within.
