nEye.ai Raises $80M Series C to Scale Optical Interconnect Chips for AI Data Centers
Santa Clara just served up a quiet flex that lands louder than most headlines. nEye.ai pulled in $80M in Series C funding, led by Sutter Hill Ventures with CapitalG, M12, Socratic Partners, and returning capital in the mix. Total raised now sits at $152M, stacking on top of a $14.5M Series A in 2023 and a $58M Series B in 2025. In a market obsessed with model size, this is a reminder that the real chokehold on progress lives deeper in the stack, right where photons start doing the heavy lifting.
Ashish Vengsarkar, CEO, and Co-Founder and CTO Tae Joon Seok, alongside Co-Founders Ming C. Wu and Kyungmok Kwon, are operating from a foundation that predates the current AI frenzy. Founded in 2020 out of UC Berkeley research, nEye.ai is the product of over a decade of work in silicon photonics and MEMS, now stepping into a phase where theory meets throughput. The transition from lab-grade innovation to production-grade infrastructure is where most companies stall. This one is accelerating.
The pitch is clean, almost disrespectfully simple. Data centers are congested, power-hungry, and increasingly constrained by latency ceilings. nEye.ai answers with its Optical Circuit Switch on a chip, pushing direct optical connections across thousands of GPUs and memory units. The numbers are not subtle: 100x smaller footprint, 1,000x lower power consumption, 10,000x faster switching, and 10x lower cost. That is not iteration, that is a structural shift in how compute moves.
This is where the startup ecosystem starts to show its tells. While attention clusters around models and applications, capital is quietly consolidating at the infrastructure layer. nEye.ai is moving from prototype into high-volume manufacturing, scaling foundry-compatible production to meet hyperscaler demand. This is no longer a research story. It is a supply chain play with real consequences for how AI systems get built and deployed.
Investor alignment sharpens the picture. Sutter Hill Ventures stepping in to lead is a signal. CapitalG and M12 maintaining position reinforces it. Add NVIDIA and Micron Ventures from earlier rounds and the pattern becomes clear. The startup ecosystem tends to reward visibility on the surface, but capital tends to concentrate where constraints get removed at scale.
The underlying shift is hard to ignore. AI is no longer just a compute problem. It is a movement problem, a bandwidth problem, and an energy problem. nEye.ai is positioning itself inside that reality, not adjacent to it. And as the startup ecosystem continues to expand around AI, the companies controlling how data actually moves may end up defining who scales and who stalls.
