As GPU deployments accelerate, moving data efficiently between thousands of chips has become just as important as increasing compute capacity. That shift is drawing investor attention toward the networking technologies that determine whether next-generation AI systems can operate at scale. Lumilens has now secured more than $700 million in Series C funding, pushing its valuation to $5.51 billion and bringing its total capital raised beyond $900 million. The financing round was co-led by Atreides Management, Bain Capital Ventures, Meritech Capital, Seligman Ventures, and Spark Capital.

The investment places Lumilens among the most highly valued companies focused on optical interconnect technology, underscoring growing confidence that networking infrastructure will become one of the defining layers of AI expansion. Rather than competing in the crowded AI chip market, Lumilens is positioning itself where data movement increasingly determines system performance. As AI clusters continue to grow, every additional processor also increases demand for faster, lower-latency optical communication across the entire infrastructure stack. Consequently, networking technologies are becoming strategic assets instead of supporting components.

AI clusters are exposing a new infrastructure constraint

Lumilens says its first scale-out networking product has already entered production AI data centers through a multi-billion-dollar customer agreement. Reaching commercial deployment within roughly two years of founding highlights how quickly demand for optical networking has intensified alongside the AI infrastructure boom. The milestone also reflects the urgency with which hyperscale operators are deploying technologies capable of supporting larger processor clusters. The industry’s attention is steadily moving beyond silicon performance toward the systems that connect thousands of processors into a unified computing platform. As AI models continue expanding in size, networking bandwidth and signal integrity increasingly shape real-world performance, making optical interconnects a critical investment area.

Building optical infrastructure across the AI networking stack

Lumilens is developing products for multiple layers of AI data center networking rather than concentrating on a single segment. Its portfolio spans near-package and co-packaged optics designed for scale-up connectivity inside AI systems, while also including 800G, 1.6T, and future generations of pluggable optical transceivers that support scale-out communication between server racks. At the center of this strategy sits the company’s LumiCore platform, which combines silicon photonics, mixed-signal integrated circuits, electrical-optical interposers, and optical system technologies into a unified architecture. The approach targets both performance improvements and manufacturing scalability as AI infrastructure grows more complex.

Furthermore, integrating these technologies under one platform could simplify deployment across hyperscale environments that demand increasingly efficient networking solutions. The newly raised capital will support expansion across engineering teams, software development, and high-volume manufacturing capabilities. Scaling production is becoming as important as product innovation, particularly as hyperscale customers seek suppliers capable of delivering networking hardware at commercial volumes without slowing infrastructure deployment.

Investors are backing AI’s next infrastructure layer

The financing signals a broader evolution in how investors view the AI infrastructure market. Capital is no longer flowing exclusively toward companies building processors or developing large AI models. Increasingly, attention is shifting toward the technologies that enable thousands of accelerators to function as a single, coordinated system inside modern data centers.

Optical networking sits at the center of that transition because AI clusters cannot scale efficiently if communication bandwidth becomes a limiting factor. Faster interconnects improve utilization, reduce latency, and help unlock the performance gains expected from increasingly expensive GPU deployments. As a result, networking infrastructure is emerging as one of the most strategically valuable segments within the broader AI ecosystem. Lumilens’ latest funding round reflects that changing investment thesis. Instead of treating optical connectivity as supporting infrastructure, investors are placing it alongside compute, memory, and power as one of the foundational technologies required to sustain the next generation of hyperscale AI data centers.