Lumilens, an optical interconnect startup founded in 2024, emerged from stealth on August 6 with more than $700 million in Series C funding co-led by Atreides Management, Bain Capital Ventures, Meritech, Seligman Ventures, and Spark Capital at a $5.51 billion valuation. Total funding now exceeds $900 million. The startup is already shipping co-packaged optics (CPO) and near-packaged optics (NPO) into production AI data centers under a multi-billion-dollar customer agreement with an unnamed major hyperscaler.
Lumilens targets the scaling bottleneck for AI clusters: copper interconnects between GPU racks have distance limits that force optical transceivers at scale-out. McKinsey projects 800G optical transceiver production could fall 40–60% short of demand through 2027, with 1.6T shortfalls persisting at 30–40%. The company manufactures in-house using proprietary interposer technology and custom robotics, positioning itself as a non-China-dependent supplier amid FCC discussions of bans on Chinese optical modules.
For architects, Lumilens represents infrastructure de-risking: GPU interconnect bandwidth now moves from a theoretical commodity to a gated supply chain. The startup's focus on non-China manufacturing and qualification paths makes it a compliance-moat play alongside pure photonics. Combined with HBM scarcity and CPU shortages, optical bottlenecks are now a primary sizing lever for data center buildout through 2027+.