EU Startup Scores $248M to Fix AI's Data Traffic Jam

Jan de Vries · 2026-04-15

A Pisa-based tech company just landed a massive $248 million investment to solve one of AI's biggest headaches: moving data fast enough. CamGraPhIC, the research arm of 2D Photonics, got the green light from the European Commission for this funding from the Italian government. They're building a new kind of optical technology that changes how data travels inside advanced computers. This isn't just any funding round. It's reportedly one of the largest single public investments ever made in an Italian DeepTech startup. That puts CamGraPhIC in a tiny group of companies worldwide trying to rethink how data moves inside future AI systems. ### What's Really Holding AI Back? "This investment goes straight to the heart of what's limiting AI today," says Ben Jensen, CEO of 2D Photonics. "Compute keeps getting faster, but data movement hasn't kept pace. Graphene-based optical technology offers a way to move vastly more data using far less power, which is exactly what the next generation of AI systems will require." Here's the thing: AI models are getting bigger and hungrier. But the real bottleneck isn't processing power anymore. It's the flow of data between chips, accelerators, and memory. Think of it like a highway system. You can have the fastest cars in the world, but if the roads are jammed, nobody's going anywhere fast. ### How CamGraPhIC's Package Stacks Up To put this in perspective, let's look at other recent deals in the photonics space: - France: Scintil Photonics raised $62 million for integrated photonics; Arago got $27 million for its photonic AI processor - UK: Lumai secured over $11 million for optical AI accelerator tech; Optalysys raised $32 million for photonic chips - Germany: Q.ANT grabbed $77 million for photonic processors Add those up and you get roughly $209 million. CamGraPhIC's $248 million package alone is bigger than all of them combined. That shows just how serious public support is getting for light-based approaches to AI infrastructure. ### The Graphene Advantage Founded in 2018 as a spinout from the Cambridge Graphene Centre, CamGraPhIC is developing graphene-based optical interconnects. These are the bridges that connect different parts of a computer system. By using graphene (a material with incredible electronic and optical properties), they can dramatically increase bandwidth density while cutting latency and energy use. Compare that to today's best silicon photonics. The difference is night and day. Graphene lets you move more data with less power and less heat. That's exactly what's needed for scaling AI and high-performance computing. ### From Research to Reality "The measure approval allows us to move quickly from innovation to execution," says Marco Romagnoli, Co-founder and CSO. "By building manufacturing capability alongside the technology, we're laying the groundwork for graphene photonics to become a practical part of future AI systems, not just a research promise." The company already raised $31 million in its Series A round back in February 2025. Now with this new funding, they'll industrialize their optical interconnect platform for AI accelerators, high-performance computing systems, and advanced data centers. ### Why This Matters for Business For anyone working in AI infrastructure or data-heavy industries, this is huge. Current systems are hitting a wall. Electrical and silicon photonic interconnects just can't keep up. They consume more power, generate more heat, and slow everything down. CamGraPhIC's approach could be the key to unlocking the next generation of AI performance. And because this is public funding under the European Commission's State Aid Framework, there's a clear focus on building real manufacturing capability. Not just research. Not just promises. Actual production. That's the kind of shift that could reshape how we build AI systems for years to come.