An automatic inverse design service for complex RF multilayer boards based on graph neural network forward solvers generates optimum layouts in a fraction time as compared to physical full-wave Maxwell solvers and thus provides a significant competitive advantage for high-speed interconnect design.
An automatic inverse design service for complex RF multilayer boards based on graph neural network forward solvers generates optimum layouts in a fraction time as compared to physical full-wave Maxwell solvers and thus provides a significant competitive advantage for high-speed interconnect design.