Nokia's AI and Cloud Sales Jump 123% as Q3 Profit Nearly Doubles

Nokia's AI and cloud sales rose 123% year-on-year in the third quarter, with €1.5 billion of new orders booked in the three months alone — the vendor's most concrete evidence yet that enterprise demand for AI infrastructure is translating into revenue rather than rhetoric.

The figures arrived alongside a near-doubling of the bottom line. Net profit for Q3 came in at €156 million, up 95% from the same period a year earlier. Group net sales reached almost €5.2 billion, a 10% increase, or 9% once currency fluctuations are stripped out — a reminder that a tenth of the headline growth was arithmetic rather than demand.

CEO Justin Hotard, who has positioned AI infrastructure at the centre of Nokia's strategy since taking the top job, said the company is structurally placed to benefit as AI workloads push outward from data centres.

"Nokia's strengths are uniquely suited to meet growing demands for AI infrastructure that is larger, more distributed and hyperconnected," Hotard said. "We are investing where we can differentiate and co-innovate with hyperscalers."

The claim of being "uniquely suited" is a competitive assertion, not a result. What Nokia did disclose is hard: a 123% growth rate and €1.5 billion in fresh AI and cloud orders. What it did not disclose is the size of the order backlog Hotard referenced, the margin profile of those AI contracts, or how much of the growth reflects one-off hyperscaler deployments rather than repeatable spend. The absence of a backlog number is notable, because vendors under pressure to prove an AI thesis usually quote one.

What did the networks business deliver?

Mobile Infrastructure, still Nokia's largest division, posted revenue of €2.9 billion, up 5% year-on-year. The company itself qualified that gain: part of it reflects income shifted forward from Q4 into Q3, meaning the underlying quarterly run-rate improvement is smaller than the percentage suggests.

The division also benefited from what Nokia called "commercial discipline" — effectively walking away from low-margin deals — and from the continued execution of cost reduction measures. Both levers lift profitability without requiring any recovery in the radio access network market, which remains sluggish across the industry.

Hotard also pointed to continued portfolio simplification, a process that has seen Nokia shed non-core assets and concentrate capital on network infrastructure and, increasingly, AI-related product lines.

Where does AI-RAN fit in?

Hotard highlighted progress on AI-RAN, the technology that folds AI computing into radio access networks, an area Nokia is developing with Nvidia. Nvidia is both Nokia's partner and an investor in the Finnish vendor, an alignment that gives the programme funding and silicon access but also ties its credibility to Nvidia's platform ambitions.

An update on the joint AI-RAN work came on 21 October at an Nvidia-hosted event, where the partners argued that momentum behind the technology is justified rather than inflated.

"AI-native networks are moving into the physical world and our AI-RAN platform is being developed to deliver the performance needed for this market," Hotard said.

The phrasing is worth noting. The platform "is being developed" — no commercial deployment volumes, customer counts or throughput figures accompanied the statement. AI-RAN remains, on this evidence, a roadmap item with strong vendor conviction behind it and limited disclosed market traction so far.

Is the Infinera integration on schedule?

The quarter's clearest execution win concerns Infinera, the optical networking company Nokia acquired in 2025. Integration is on track to complete by the end of 2026, a full year ahead of the original plan.

Hotard said the accelerated timeline has produced €200 million in net synergies — roughly two-thirds of the quarter's entire net profit, which puts the value of the faster integration into perspective.

Nokia now enters the final quarter with profit momentum, a fast-growing AI order book and an optical integration running ahead of plan. Whether the AI growth rate survives contact with hyperscaler budget cycles will be the test of the fourth quarter and beyond.