A 23.5–32.5GHz, 17dBm PSAT and 37.5% PAE Power Amplifier Synthesized Using an Automated Design Methodology

The design of power amplifiers (PAs) in the millimeter-wave (mm-Wave) regime is remarkably complex. In order to meet the linearity and efficiency requirements for 5G or 6G operation, designers have been struggling to efficiently design such PAs in a reasonable time. Hence, in the past few years, a growing interest has emerged in developing optimization-based methodologies which can automatize the design of mm-Wave circuits. In this paper, a fully-automated, process, voltage and temperature (PVT), and layout-aware optimization-based design strategy is applied to the design of a mm-Wave PA. It accurately models the transformers in the PA using machine-learning techniques and exploits a multi-objective optimization algorithm, achieving a Pareto optimal front (POF) containing hundreds of PAs over the landscape of the performance trade-offs. By leveraging the layout in the loop, it outputs a set of design solutions that have their GDSII readily available and ready to fabricate without the need for further designer intervention.