Analog synthesis 3.0: AI/ML to synthesize and test analog ICs: hope or hype ?
Georges E. Gielen · 2023
Analog/mixed-signal integrated circuits remain indispensable in electronics applications that interface the physical with the cyber world. But whereas digital circuits are largely synthesized through EDA software, surprisingly, the design of analog circuits in industry is mainly still handcrafted, resulting in low design productivity with long and error-prone design cycles and high development costs. A showstopper is the large user interaction needed to steer current circuit and layout optimization tools towards acceptable solutions. This keynote will explore the high potential of using advanced machine learning (ML) techniques to overcome these issues and even to self-learn and then exploit design expertise and constraints from available completed designs. Such innovations may finally enable to fully automate analog circuit design without human designer in the loop. In addition, the presentation will also cover the use of AI techniques in analog test program development, focused on boosting the analog fault coverage as needed in safety-critical applications.