Variation-Aware Methods and Models for Silicon Photonic Design-for-Manufacturability

Methods for the generation and application of process variation-aware compact models for silicon photonics are presented. Three categories of process variations are considered. First, systematic variations, including die-to-die and spatial variations, are addressed with parameterized variation-aware compact models, and demonstrated for coupled resonator optical waveguide variation impact and yield analysis. Second, methods for random variation are presented, with ensemble and adjoint-based variation modeling methods. Third, particle defects are addressed with an extension to adjoint methods, enabling efficient analysis and identification of photonic component areas that are sensitive to defects. Finally, methods for extraction of variation models using variation test chip design and measurement are presented. Together, these methods are key components toward design-for-manufacturability approaches to achieve high yield and high performance in photonic integrated circuit design.