nEye Systems
Series B - $58M Raised
Some startups whisper. Others echo. But every so often, one hits the floor like a dropped amp, rewiring the entire sound system on the way down. That’s nEye Systems, a photonics disruptor born in UC Berkeley labs, now stacking Series B chips to the tune of $58 million. CapitalG led the charge with M12, Micron Ventures, NVIDIA, and Socratic Partners riding shotgun.
Let’s rewind. In 2020, a crew of visionaries turned lab-grade lightwaves into enterprise-grade infrastructure. Professor Ming Wu, a photonics pioneer with more patents than your favorite startup has Slack channels, teamed up with MEMS specialist Tae Joon Seok and founding engineer Xiaosheng Zhang, who could probably design an optical switch in his sleep. Together, they spun out nEye Systems, bringing decades of research from Berkeley’s labs into the hyperscale game.
Their flagship tech, SuperSwitch, isn’t just a clever name; it’s a reality-bender. Picture a wafer-scale optical circuit switch that connects thousands of GPUs and memory units without choking power or dragging latency. It’s 100x smaller, sips 1,000x less juice, and costs 10x less than the nearest legacy dinosaur pretending to be future-ready. It’s software-reconfigurable, high-radix, and whisper-quiet in an industry addicted to fan noise and cable spaghetti.
This round isn’t a vanity boost. It’s the build phase. CapitalG’s James Luo now sits on the board, and Sheppard Mullin inked the deal. Meanwhile, engineering is scaling, production samples are on deck for 2026, and a 2-to-10 person team is punching way above its weight class. Think less “stealth mode,” more “subsonic disruption.” Hyperscale data centers are already looking sideways at their own power bills, wondering if they just met their next infrastructure upgrade.
What makes this one worth watching? Simple. nEye didn’t chase a trend, they solved a bottleneck the size of a city block. AI and ML workloads aren’t just demanding; they’re insatiable. And while everyone else is arguing over GPU count, nEye’s team asked the real question: What’s connecting them? Their answer? Optics that scale like code and move data at the speed of light, with a power footprint so small it practically moonwalks past traditional switching.