We are thrilled to announce that CapitalG is leading a $58M Series B in nEye Systems, an optical circuit switch platform revolutionizing data center infrastructure. It’s a privilege to partner with Professor Ming Wu and the exceptional nEye team as they tackle one of the most pressing constraints in modern computing: more efficient data center networking.
The Data Traffic Jam Stifling AI’s Potential
The artificial intelligence revolution is upon us, creating efficiencies in the ways we live and work and reshaping industries and daily life at an astonishing pace. Yet, beneath the surface of what may seem like a blistering pace of AI advancement, a fundamental challenge bottlenecks the infrastructure powering this transformation: data center networking.
Processors have become incredibly powerful, yet electrical switches face significant challenges in the effort to keep up, leading to bandwidth shortfalls and excessive power consumption. The result: congestion, spiraling costs, and underutilized processing capacity. This is more than a technical hiccup; it directly impacts AI’s future. Imagine trying to propel a fleet of supercars through narrow, congested city streets with ineffective traffic lights and intersections. That’s analogous to the situation inside today's data centers.
While processors (the engines of AI) have become phenomenally powerful, the traditional electrical switches directing massive flows of data between them simply haven't kept pace. They struggle under the load, generating excessive heat, consuming vast amounts of power, and creating bandwidth shortfalls. The result is a critical traffic jam for data, resulting in severe network congestion, sky high operational costs, and incredibly powerful, expensive processors sitting idle, starved of data to compute.
This data traffic jam is a direct impediment to AI's trajectory. As AI models grow increasingly complex, the volume of data that must move seamlessly and instantly between thousands of compute nodes explodes. Without fundamentally new network architectures, data centers risk hitting performance ceilings that will hobble AI’s potential and delay breakthroughs. Major cloud providers are already exploring optically reconfigurable supercomputers, signaling the industry’s move toward high-bandwidth, low-power solutions.
The nEye Approach: Let There Be Light-Speed Data
Enter nEye Systems. Born from over a decade of pioneering research in Professor Ming Wu’s acclaimed UC Berkeley lab, nEye offers a groundbreaking Optical Circuit Switch (OCS) designed to clear these data center bottlenecks. Instead of forcing data through power-hungry electrical pathways, nEye uses the speed and efficiency of light. Their unique silicon photonic switch delivers unparalleled performance combined with radically lower power consumption. In fact, its performance is:
100x smaller than alternative optical switching solutions
1,000x lower in power consumption
10,000x faster at switching
10x lower cost
Professor Wu is a world-renowned luminary in the field of silicon photonics and MEMS (micro-electromechanical systems). With over 600 scholarly papers and 53 patents, he is widely considered to be a pioneer of optical circuit switching. Professor Wu now leads the incredibly talented nEye team, translating deep academic innovation into a commercial solution ready to unblock the constraints throttling AI performance. Their wafer-scale OCS technology not only alleviates current network pressures, but also enables entirely new, flexible, and resilient network designs optimized for the most demanding high-performance computing and AI/ML clusters.
CapitalG’s Investment in nEye
At CapitalG, we invest in transformative companies poised to reshape industries. nEye perfectly fits this vision. We believe their OCS technology represents a fundamental architectural shift necessary for the next era of computing. We are incredibly proud to lead their $58 million Series B round, alongside Microsoft’s M12, Micron Ventures, NVIDIA, and Socratic Partners.
We firmly believe that efficient, high-bandwidth networking is the critical substrate upon which future AI progress will be built. We look forward to a close partnership with Professor Wu and the entire nEye team as they help data centers worldwide overcome these critical networking challenges – and, in doing so, unlock AI’s true potential.