Introduction to PreFab Photonics: AI-Powered Virtual Nanofabrication for Your Next Tape-Out
Event Details
Photonic devices demand nanometer precision. At that scale, manufacturing variations are
unavoidable—but standard design tools assume ideal geometry.
PreFab addresses this challenge. It is a Python-based tool that uses deep learning to predict
real manufacturing outcomes from design layouts—before tape-out. With built-in uncertainty
quantification, designers can work with realistic margins and increase their chances of hitting
design specs.
If you are designing custom photonic components—rings, gratings, photonic crystals, or
inverse-designed structures—join this webinar to learn how PreFab can help.
This webinar will cover:
- An introduction to PreFab and how it predicts fabrication outcomes from design files
- An overview of the fabrication predictor and corrector models, and where they have the most impact
- A live demonstration of the workflow for setting up, running, and analyzing predictions
- How to validate PreFab predicted fabrication outputs using your existing FDTD simulation tools
Register to see how PreFab can help you reduce iteration cycles, hit specs faster, and achieve
more reliable results on your next tape-out.
PreFab is available to CMC academic subscribers. Currently supports Applied Nanotools
(ANT), with additional foundry models available upon request.
Tue, Feb 24, 2026 1:00 PM – 2:00 PM (UTC-05:00) Eastern Time (US & Canada)
We look forward to having you attend the event!
