AI-Enabled Discovery of New RF/mm-Wave Architectures and Synthesis of End-to-End RFICs

Traditionally, chip-scale RF system design has been the domain of the expert, dominated by rules of thumb and trial-and-error techniques. Designing these ICs, that form the bedrock of the wireless networks, is complex, time-consuming, requires years of expertise, and therefore, can be very expensive. Historically, the process of RF IC design has relied on intuition-based approaches with standard templates that are subsequently optimized, time-consuming parameter sweeps, or ad-hoc population-based metaheuristic optimization methods. There is no reason to believe that this approach is optimal in any sense. This talk will discuss how inverse design with AI-based approaches can open a new design space and allow rapid designs on demand. We will demonstrate such algorithmic synthesis approaches with designed and fabricated mm-wave and sub-THz ICs.