Agentic data science with the Data Agent Kit
cloud.google.com·7h ago
AI Research
arXiv CS.AI·Jul 29·5 min read
Physics-Informed Neural Operator for Warm-Starting Background-Decomposed and Preconditioned PSFD: Enabling Scalable 3-D EUV Mask Simulation
TL;DR
A physics-informed neural operator (PINO) was developed to enhance electromagnetic scattering simulations in extreme ultraviolet (EUV) lithography. This model significantly reduces computational costs while achieving high accuracy in predicting scattered intensity.
✦ Why It Matters
Implement the PINO model to enhance your EUV lithography simulations and reduce computational costs.
Key Takeaways
- Utilize the PINO for efficient EM scattering simulations.
- Train on diverse mask designs to improve model accuracy.
- Leverage warm-start initialization to speed up solvers.
- Reduce computational costs by minimizing domain size.
How It Works
The PINO is structured with a two-dimensional lateral branch and a one-dimensional axial branch, allowing it to efficiently model complex EM interactions without relying on finite-order approximations. By training on a large dataset of mask designs, the model learns to predict scattering behavior accurately, which is crucial for optimizing lithography processes.