Graphene optical interconnect technology is moving from research labs into real manufacturing, and 2D Photonics just secured the funding to make that happen at scale. The bottleneck in modern AI systems is no longer processing power. It’s data movement. As AI models scale and hardware packs more compute into smaller spaces, moving data between chips, memory, and accelerators becomes slower, hotter, and more power hungry. That’s the problem 2D Photonics, through its subsidiary CamGraPhIC, is now funded to solve.
On April 15, 2026, the European Commission approved a €211 million Italian State aid grant for CamGraPhIC, the wholly owned research arm of 2D Photonics. The funding will go toward industrializing graphene-based optical interconnect technology for AI accelerators, high performance computing systems, and data centers. This makes it one of the largest single public investments ever made in an Italian DeepTech startup.
The Core Technology:
CamGraPhIC is developing graphene-based Optical I/O, a class of optical technology that replaces conventional electrical and silicon photonic interconnects. The goal is to carry significantly more data at lower latency while consuming far less power than current silicon photonics solutions.
Graphene, the material at the center of this work, has strong electronic and optical properties that make it well suited for high speed data transmission at high bandwidth density. Rather than making incremental improvements to existing approaches, CamGraPhIC is building a different class of optical link altogether, one designed to sit closer to the processor itself.
Breaking Down the Grant:
Ben Jensen, CEO of 2D Photonics, put it simply: compute keeps getting faster, but data movement hasn’t kept pace. The €211 million is not research funding in the traditional sense. It’s industrialization capital, meaning the company will use it to build a manufacturing facility near Milan and develop a direct path from lab scale production to high volume foundry deployment.
The pilot manufacturing line is scheduled to become operational in 2028. At that point, CamGraPhIC will begin early engagement with system integrators and computing platform providers. The project is also expected to create over 150 jobs in photonics engineering, materials science, and semiconductor manufacturing.
From Research to Production:
Marco Romagnoli, Co-founder and CSO of CamGraPhIC, noted that this approval allows the company to move quickly from innovation to execution. The construction of the Milan area facility is a key part of that. It’s designed to handle device qualification, early manufacturing runs, and technology transfer to larger foundry processes.
This approach, building manufacturing capability alongside the technology, means that the graphene photonics work being done at CamGraPhIC is intended to become a deployable product, not a long horizon research project. For anyone tracking AI infrastructure investments in Europe, this is a meaningful step toward domestic DeepTech production capability.
2D Photonics and CamGraPhIC:
2D Photonics raised €25 million in its Series A round in February 2025, with backing from CDP Venture Capital, NATO Innovation Fund, Join Capital, Sony Innovation Fund, Bosch Ventures, Indaco Ventures, and Frontier IP Group. The group focuses on photonic circuits for telecom, datacom, and high performance computing applications.
CamGraPhIC was founded by Marco Romagnoli and Professor Andrea Ferrari. As a subsidiary of 2D Photonics, it concentrates specifically on optical interconnect solutions for next generation compute. The €211 million grant puts the company in a small group globally working on this class of technology.
The Larger AI Infrastructure Picture:
AI infrastructure investment is accelerating across hardware, software, and physical facilities. Optical interconnects are increasingly recognized as a key layer in that stack, particularly as the power and thermal constraints of large scale GPU clusters become more acute.
CamGraPhIC’s Optical I/O platform targets exactly those constraints. Founders and operators building or evaluating AI infrastructure should note that this is an area where the underlying physics of silicon photonics are driving demand for alternatives. Graphene based approaches are one of the more credible paths being developed at scale right now.
For a broader view of what AI infrastructure funding looks like across Europe right now, the recent wave of DeepTech grants offers useful context on where public capital is going alongside private venture dollars.
