Neurophos has raised $110M in a funding round led by Gates Frontier to advance optical computing technology. Based in Austin, Texas, the company uses silicon photonics circuits and meta-materials to perform optical domain computations, addressing the massive energy consumption challenges faced by modern data centers.
Traditional computing relies heavily on electronic circuits, moving digital bits through transistors, switches, and memory while adhering to Moore’s Law. However, data centers encounter massive quantities of optical data that traditionally require conversion into the electrical domain—shifting from photons to bits—to handle switching and compute operations for tasks like online search and artificial intelligence training and learning.
This conversion from optics to electronics and back to optics, known as OEO conversion, is highly lossy and significantly increases both energy and water consumption in data centers. A hyperscale data center consumes between 100MW and 1GW of power, with roughly 50 percent of that energy dedicated to compute operations.
Austin-Based Neurophos Raises $110M Led by Gates Frontier
To combat power walls that constrain traditional GPUs, Neurophos is utilizing silicon photonics and meta-materials to perform computations directly in the optical domain. Headquartered in Austin, Texas, the company secured $110M in recent funding led by Gates Frontier, with participation from M12 (Microsoft’s Venture Fund), Carbon Direct Capital, Aramco Ventures, Bosch Ventures, Tectonic Ventures, Space Capital, and others. This capital injection enabled the company to scale its workforce by 4X in 2026.
Moore’s Law is slowing, but AI can’t afford to wait. Our breakthrough in photonics unlocks an entirely new dimension of scaling, by packing massive optical parallelism on a single chip.
Dr. Patrick Bowen, CEO and Co-Founder of Neurophos
Dr. Patrick Bowen noted that the physics-level shift improves both efficiency and raw speed as systems scale up. By breaking free from conventional power constraints, optical compute alters the amount of artificial intelligence that data centers can deliver relative to power, space, and monetary investments.
Optical Circuit Switching and the Evolution of Optical Computing
Optical Circuit Switching, or OCS, minimizes expensive OEO conversion by utilizing silicon photonics and meta-materials to switch optical beams. While optical computing concepts originated in 1960 when researchers found that lenses and optical filters could perform Fourier transforms, the field progressed slowly. Spatial light modulators emerged in the 1970s to create programmable patterns of light, followed by academic research spanning from 1980 to 2000 that remained largely confined to laboratories while electronic computing advanced rapidly.
A critical breakthrough occurred when spatial light modulator technology was adapted to execute vector matrix multiplication. Between 2000 and 2025, optical semiconductors, fiber-optics, free-space optics, and silicon photonics matured, fueling growth for multi-billion optics players such as Lumentum, Coherent, Corning, and IPG Photonics.
Inside optical tensor cores, addition is performed by focusing light reflected from each column of a grid onto a single photodetector, adding the light into a single element of the output. Because modulators retain their matrix while thousands of input vectors pass through, loading energy is distributed across all of them. These operations occur on optical chips featuring tiny waveguides etched into silicon, SiN, or InP to move optical beams, alongside integrated laser sources, optical detectors, and modulators that alter properties like intensity, phase, polarization, and wavelength.
According to the Yole Group, the first OCUs are expected to deploy in 2028, with the market expected to grow to ~$2.7B (1 million units) by 2034. This technology targets the broader data center compute market currently estimated at ~$1T/year today.
