Lumilens, a connectivity platform purpose-built for AI infrastructure, has closed more than $700 million in new funding, the company announced, in a round co-led by Atreides Management, Bain Capital Ventures, Meritech, Seligman Ventures, and Spark Capital. The financing values the company at $5.51 billion and lifts its total capital raised to over $900 million, underscoring investor conviction that optical interconnect technology has become a critical chokepoint in the global AI buildout.
The round drew participation from a broad slate of institutional and strategic investors, including Addition, Aiconic, Alkeon, Harbourvest, J.P. Morgan Private Capital, Lingotto, Mayfield, MVP Ventures, Peak XV, Qualcomm Ventures, Redpoint Ventures, Seifdune, Thomvest Ventures, and EDBI, the investment arm of Singapore's EDB and Enterprise Singapore. The capital will be directed toward scaling the company's silicon, systems, software, process engineering, and high-volume manufacturing operations.
Lumilens was founded in early 2024 by a team that set out to address the connectivity challenges that the largest AI clusters would inevitably face — problems that a leading hyperscaler was already confronting as it planned the scale of clusters its AI roadmap would require. Within roughly two years, the company's initial scale-out product completed qualification and began shipping into that hyperscaler's production AI data centers under a multi-billion-dollar agreement, a pace that has become a key proof point for investors.
The company designs and manufactures optical interconnects spanning both network layers inside an AI data center: the scale-up network that directly connects GPUs within a single compute system, and the scale-out fabric that links racks and clusters together. Its portfolio includes scale-up silicon and systems based on near-package optics (NPO) and co-packaged optics (CPO), which bring optical I/O directly to the GPU and are designed to enable thousands — and eventually tens of thousands — of processors to operate as a single, tightly coupled domain. On the scale-out side, Lumilens produces pluggable optical transceivers at 800G, 1.6T, and beyond, replacing copper links across racks and rows to carry more data at lower power for the largest AI fabrics.
The funding comes as industry attention shifts from raw GPU supply toward the networking layer that determines how effectively those processors can work together. "The constraint on AI has shifted from how many GPUs you can buy to how many you can connect," said Ankur Singla, founder and CEO of Lumilens. "Whether they're training frontier models or serving them to hundreds of millions of users, hyperscalers told us they need the same two things: far more optical capacity for the networks they run today, and a path to directly connecting thousands of GPUs together into a single cluster. We built Lumilens to deliver both — and to manufacture at the volume, quality, and pace this buildout demands."
Investors echoed that view. "Over the past several years, the conversation around AI infrastructure has centered on access to compute," said Umesh Padval, managing partner at Seligman Ventures. "Increasingly, the constraint today is shifting to connectivity and how efficiently thousands of GPUs can operate as a single system." Gavin Baker, managing partner at Atreides Management, noted that Lumilens went from founding to a qualified product deployed inside a hyperscaler's production data centers within two years, and is now building the manufacturing engine to scale with demand while bringing much-needed diversification to AI supply chains. Navin Chaddha, managing partner at Mayfield, said copper has hit a wall the industry cannot engineer around, adding that Mayfield partnered with Singla for the third time, led Lumilens' seed round, and has participated in every round since.
To pursue what the company estimates is a $100 billion-plus market opportunity for photonic interconnects, Lumilens has assembled a leadership team with deep expertise across photonics and large-scale networking, drawing executives and engineers from Cisco, Juniper Networks, Meta, Marvell, Lumentum, and Coherent. The company's ability to move from design to high-volume production — treating manufacturing as a product in its own right, from chip assembly through volume output — is central to its pitch as AI operators race to scale clusters and diversify supply chains.
Lumilens Raises Over $700M at $5.51B Valuation to Break AI Data Center Connectivity Bottlenecks
By: IDCNOVARegion: Southeast Asia