About Neurophos
Neurophos is developing an optical processing unit, a chip that performs the matrix multiplications at the heart of AI inference using light instead of electric current. Its approach uses metasurface modulators, micron-scale structures built from engineered metamaterials, which the company says can be made in standard CMOS factories. It positions the chip as an alternative to GPUs for data center inference, though performance figures are company claims. Neurophos is based in Austin, Texas, and spun out of Duke University and the Metacept incubator.
Neurophos was founded in 2020 and is based in Austin, United States. It is listed under Photonic Computing, AI Accelerators. Reported funding is $110M Series A (Jan 2026, Gates Frontier-led; total $118M).
Compute Technologies
- Photonics
- Neuromorphic
What Neurophos works on
Photonic Computing
Photonic computing uses light instead of electricity to carry or process information. Photons have no charge, so light in a waveguide does not cause the resistive heating that current in a wire does, though lasers and modulators still use power, and many wavelengths can share one waveguide. The industry has two quite different parts: moving data with light, and computing with light.
AI Accelerators
An AI accelerator is a chip, or a part of a chip, designed to run machine learning faster or with less energy than a general-purpose processor. The category includes GPUs, TPUs, neural processing units (NPUs) and chips from many startups.
Also in United States
Luminous Computing, Broadcom, Marvell Technology, Oxmiq Labs, Lemurian Labs
