Nokia says its work with NVIDIA has carried voice, mobile data and video analytics simultaneously on Elisa’s live 5G network in Helsinki. The October 21 update describes a concrete test of shared AI and radio infrastructure, while placing wider commercial services in 2027. Nokia AI-RAN commercialisation update.
The distinction matters because a radio network has service obligations even when an additional AI workload competes for computing resources. Demonstrating mixed workloads under selected conditions is a useful step; it does not establish performance for every traffic pattern or an already general-purpose commercial service.
Resource sharing needs an operating rule
The engineering problem is deciding what happens when radio demand and AI demand rise together. Operators need predictable priorities, monitoring and recovery behaviour. Our assessment is that useful utilisation gains require the system to protect the connectivity service customers already buy.
Nokia also describes work with other operators. Results from a laboratory test at one partner should not be assigned to the Helsinki live-network work, and planned trials should not be counted as completed deployments.
Follow commercial service evidence
The next useful disclosures would state load conditions, service continuity and the capacity left for AI after radio requirements are met. If the approach performs consistently, it could broaden the work supported by operator infrastructure. The reported demonstration provides evidence to investigate that possibility, rather than a guarantee of network-wide economics.
Image: Hannes Grobe / CC BY-SA 4.0. Reuse terms. Original retained; responsive display may crop the frame.
Sources & evidence
Sources describe the announcements and evidence available when this analysis was published. Company claims are attributed; plans and trials are not treated as sustained operation.
