Layout-Aware Variability Analysis, Yield Prediction, and Optimization in Photonic Integrated Circuits

We present a simulation framework for evaluating the effect of location-dependent variability in photonic integrated circuits. The framework combines a fast circuit simulator with circuit layout information and wafer maps of waveguide width and layer thickness variations to estimate the statistics of the circuit performance through Monte Carlo simulations. We illustrate this with ring resonator filters, a design sweep of Mach-Zehnder lattice filters, and the tolerance optimization of a Mach-Zehnder interferometer, and show how variability aware design can be essential for future photonic circuit design, especially in a fabless ecosystem where details of the foundry processes are not available to the designers.