Lumilens Exits Stealth With $700M Series C to Rewire AI Data Centers With Light

San Jose silicon-photonics startup Lumilens emerged from stealth on August 6, 2026 with over $700 million in Series C funding at a $5.51 billion valuation, backed by Atreides, Bain Capital Ventures, Meritech, Seligman, Spark Capital and Mayfield.

Lumilens Exits Stealth With $700M Series C to Rewire AI Data Centers With Light

Silicon-photonics startup Lumilens exited stealth on August 6, 2026 with more than $700 million in Series C funding, valuing the company at $5.51 billion and lifting total capital raised past $900 million. The round was co-led by Atreides Management, Bain Capital Ventures, Meritech, Seligman Ventures and Spark Capital, with participation from Addition, Alkeon, HarbourVest, J.P. Morgan Private Capital, Mayfield, MVP Ventures, Qualcomm Ventures, Peak XV, Redpoint Ventures, Seifdune, Thomvest Ventures, EDBI and Lingotto.

Two years from founding to hyperscaler production

Founded in early 2024 by CEO Ankur Singla, Lumilens designed a full portfolio of near-package optics (NPO), co-packaged optics (CPO) and 800G/1.6T pluggable transceivers built on its in-house LumiCore platform. Its first scale-out product finished qualification and is already shipping into a hyperscaler's production AI data centers under what the company calls a multi-billion-dollar customer agreement.

Why connectivity is the new AI bottleneck

The pitch: buying GPUs is no longer the hard part of scaling AI — connecting them is. A single 400,000-GPU data center requires more than 2.4 million optical transceivers and five million fiber strands, and McKinsey projects 800G optics could fall 40–60% short of demand through 2027. Inside the rack, copper links between GPUs cap out around 1.5 meters at AI data rates, forcing hyperscalers to move to photonics to scale up beyond a single cabinet. Lumilens builds for both problems — the scale-up fabric inside the rack and the scale-out fabric across rows.

Lumilens logo

Investor voices

"The constraint on AI has shifted from how many GPUs you can buy to how many you can connect," Singla said. Atreides managing partner Gavin Baker called the two-year path from founding to qualified hyperscaler product "a formidable combination" of speed and scale, while Mayfield’s Navin Chaddha — who has now backed Singla three times — said the firm has "never seen execution and growth like this" in 56 years of investing. The capital will expand Lumilens' silicon, systems, software, process engineering and high-volume manufacturing operations.

Related coverage on The Robotics Media: U.S. Commerce Department lines up $874M in AI chip letters of intent, Anthropic’s $10B Volta compute deal, and Hadrian’s $1.37B Series D for AI factories.

Reporting based on coverage from Lumilens, SiliconANGLE, Dealroom, TechFundingNews and Finsmes.

Category: Funding & Investments

Tags: data centers Series C Funding AI Infrastructure Photonics GPU Computing AI Chips

Related Articles