$NOK

PON (Passive Optical Network) is a telecommunications architecture that uses optical splitters, which do not require electrical power, to distribute a single fiber optic line across multiple devices or endpoints. This structure offers major advantages such as lower power consumption, simpler infrastructure, and more efficient network design.
The solution Nokia announced today is a PON-based out of band management system developed for data centers. According to the company, this technology can reduce the number of active switches by 90%, deliver more than 50% energy savings, and simplify operational workload by 80%. Nokia also stated that its Aurelis solution is already being used in more than 700 mission-critical enterprise networks. That detail matters, because this is not just a theoretical vision. It is about adapting a field proven technology to the data center environment.
The real critical point is this: integrating PON technology into AI data centers could create a massive and high growth new Total Addressable Market, or TAM, for Nokia. The company has recently been clearly repositioning its strategy around the AI supercycle. In its 2025 annual report, it defined AI & Cloud as one of its five strategic priorities, highlighted that the addressable market in this segment is growing 28% annually, and disclosed that it generated €2.4B in orders from AI & Cloud customers during 2025. That shows this shift is not accidental, but strategic.
For that reason, this news means more than an ordinary product announcement. Nokia no longer wants to be only an optical transport or traditional telecom backhaul player within the data center ecosystem. The company is also trying to enter the connectivity and management layer inside the data center itself, in order to capture a larger share of the AI infrastructure spending cycle. If this approach gains commercial traction, Nokia's growth story may begin to be driven not only by traditional telecom infrastructure, but also by the next generation network architecture of AI data centers.