AI Designed Photonic Chips for AI Infrastructure

We harness AI to design photonic and RF chips for datacenter interconnect, consumer electronics and quantum computing.

Product

Software

Generative AI meets Maxwell's Equations! We created a fully differentiable and GPU-accelerated finite difference time domain (FDTD) engine for electromagnetic simulation and inverse design. Our AI can design photonic and wireless integrated circuits, meta-surfaces, and even antennas! See our product documentation for details. Follow us on LinkedIn for updates!

Components

Our AI can design perfectly vertical grating couplers, edge couplers, wavelength domain multiplexers and more! All with unprecedented performance and compact footprint.

Chips

We’re designing silicon photonic chips for fiber interconnect in datacenter infrastructure. Both pluggable transceivers and co-packaged optic (CPO) modules.

Team

Founder: Paul Shen | Stanford MS BS
Adviser: Steven G Johnson | MIT Faculty