nanoPMC: a GPU-accelerated particle Monte Carlo simulator for feature-scale semiconductor process modeling

We present nanoPMC, a GPU-accelerated, Python-based, feature-scale Particle Monte Carlo (PMC) process simulator developed for modeling etching and deposition processes in semiconductor fabrication. The simulator is designed with modularity and flexibility, incorporating distinct modules for structure generation, particle generation, transport, surface interaction, and animation. Leveraging CUDA-enabled GPU parallelism, nanoPMC achieves high-speed computation while maintaining accuracy. The simulator has been validated for trench etching profiles under varying input distributions and applied to key Gate-All-Around (GAA) FET fabrication steps, including Inner Spacer Etching, Sheet Release, and Gate Metal Deposition. A GUI-based web application is under development to improve accessibility and extend the tool’s utility for process simulation and optimization.