- Lumilens raises more than $700 million in Series C funding, valuing the AI infrastructure startup at $5.51 billion.
- The optical interconnect startup is already shipping into a hyperscaler’s live data centres.
- Repeat founder Ankur Singla says connectivity, not chips, is AI infrastructure’s next constraint.
Lumilens has emerged from stealth with more than $700 million in fresh funding, valuing the AI infrastructure startup at $5.51 billion and pushing its total capital raised above $900 million.
The Series C was co-led by Atreides Management, Bain Capital Ventures, Meritech, Seligman Ventures, and Spark Capital, with more than a dozen other investors joining, including Qualcomm Ventures and J.P. Morgan Private Capital.
Two years after its founding, Lumilens is already shipping its first optical interconnect products into a hyperscaler’s production data centres, under what the company describes as a multi-billion-dollar customer agreement.
“The constraint on AI has shifted from how many GPUs you can buy to how many you can connect. Hyperscalers need far more optical capacity and a path to directly connecting thousands of GPUs together into a single cluster. We built Lumilens to deliver both, and manufacture them at the scale this buildout demands,” said Ankur Singla, founder and chief executive of Lumilens.
Why connectivity is AI’s new bottleneck
For the past three years, the AI industry’s defining scarcity was the GPU itself. That is starting to change. As hyperscalers rack up ever-larger clusters, the copper wiring that moves data between chips is running out of headroom, and the bottleneck is shifting from compute to connectivity.
Citigroup estimates AI infrastructure spending will exceed $2.8 trillion by 2029, and a growing share of that is being funnelled into the optics needed to keep pace: Nvidia alone has committed $2 billion toLumentum for silicon photonics manufacturing.
Lumilens is entering a field that has become one of the most heavily capitalised corners of AI infrastructure in the past year. Lightmatter raised a $400M in Series D funding in October 2024 which brought the company’s valuation to $4.4B. Ayar Labs pulled in $500 million from Nvidia and AMD at a $3.75 billion valuation in March 2026. Celestial AI closed a $750 million Series D in June 2026 at a $6 billion valuation, backed by Fidelity and BlackRock.
Smaller players are chasing the same problem from different angles: Spain’s iPronics, Zurich’s Aylight, and Italy’s CamGraPhIC. Lumilens says it is already shipping.
A repeat networking founder’s fourth act
Singla has done this before. He founded Contrail Systems, a software-defined networking company that Juniper Networks acquired in 2012 for roughly $176 million, and then built Volterra, an edge-cloud platform that F5 bought for close to $500 million in January 2021.
Between Volterra and Lumilens, he also started Exaforce, an AI security platform now valued at roughly $725 million, where he remains executive chairman. Lumilens is his fourth infrastructure company.
He is joined at Lumilens by co-founder and chief technology officer Ted Schmidt, a former distinguished engineer at Juniper Networks who worked on the company’s silicon photonics programme after its 2016 acquisition of Aurrion. Schmidt has since held senior roles at Lumentum and Effect Photonics
How Lumilens plans to out-build the field
The startup, founded in 2024, operates by building what it calls the LumiCore platform: a stack that combines silicon photonics, mixed-signal integrated circuits, optical systems, and its own manufacturing processes, designed to take products from initial design to customer qualification in months rather than years.
Lumilens builds for both scale-up networks, which connect GPUs within a single system, and scale-out networks, which link racks across a data centre, with a long-term roadmap aimed at binding tens of thousands of GPUs into one tightly coupled cluster.
While rivals such as Ayar Labs and Lightmatter have focused on chip-to-chip optical links, Lumilens is aiming to own both layers of the stack at once, alongside the manufacturing capacity to produce at hyperscaler volume.
Its leadership team, which includes Ritesh Kapahi as vice president and general manager for its India operations, draws engineers from Cisco, Juniper Networks, Meta, Marvell, Lumentum, and Coherent.
“We are excited to partner with Ankur Singla, a proven repeat founder who has assembled a world-class team spanning silicon, optics, software, systems, supply chain and manufacturing expertise to address one of the most important challenges in modern data centres. Lumilens’ early hyperscaler adoption and multi-billions in customer commitments just two years after its founding reflect both the scale of the opportunity and the team’s ability to execute,” said Umesh Padval, managing partner at Seligman Ventures.
Gavin Baker, managing partner at Atreides Management, zeroed in on the timeline: “Focusing on that core AI challenge, Lumilens went from its founding to a qualified product deployed inside a hyperscaler’s production data centres within two years, and is building the manufacturing engine to scale with demand while bringing much-needed diversification to AI supply chains. Speed and scale are a formidable combination, and we’re thrilled to lead this round.”
Mayfield has backed all three of Singla’s prior companies, and managing partner Navin Chaddha wasn’t shy about saying why: “Connecting GPUs is the bottleneck now, and copper has hit a wall the industry can’t engineer around. That’s why we partnered with Ankur Singla for the third time to help build Lumilens from inception, led their seed round, and have participated in every round since.”
James Kuklinski, general partner at Spark Capital, called it “a rare combination of deep expertise in silicon photonics, networking systems and high-volume manufacturing,” adding that the firm was “proud to partner with Ankur Singla and the world-class team at Lumilens.”
Lumilens will use the new funding to expand its silicon, software, systems engineering, and high-volume manufacturing operations to keep pace with hyperscaler demand. Headcount plans were not disclosed. The Series C brings its total funding to more than $900 million
