1. What did Nokia report for the third quarter?
Nokia reported third-quarter 2026 results on October 22, with net sales of €5.154 billion, about 10% higher than a year earlier, and a 123% jump in sales to AI and cloud customers. Orders from that customer group climbed 118% to €1.5 billion, and the comparable operating margin reached 11.7%. Chief Executive Justin Hotard left full-year profit guidance unchanged.
The Finnish company has long been known for mobile networks, and for many years investors judged it almost entirely on how much telecom operators chose to spend on radio equipment. The latest quarter shows a different engine at work. The strongest numbers came from selling to the operators of large data centers, the companies that are building the computing capacity behind artificial intelligence services.
The result also came in ahead of what analysts had been expecting on profit, which is why the update was received as a beat rather than a simple confirmation of the existing plan. The figures are reported by the company itself, and most of the forward-looking statements that accompany them are management guidance rather than independent forecasts.
2. The figures behind the quarter
Net sales of €5.154 billion represent growth of roughly 10% compared with the third quarter of 2025. For a company of Nokia's size, double-digit expansion is not routine, and it reflects the fact that a fast-growing business line is now large enough to move the group total.
That business line is the one serving AI and cloud customers. Its sales rose 123% year over year, which means the segment more than doubled in twelve months. When a part of a company grows at that pace, even a modest share of revenue can account for a large portion of the overall increase, and the data suggests that is what happened here.
Orders tell a similar story. Order intake from AI and cloud customers rose 118% to €1.5 billion during the quarter. Sales describe what was delivered and recognized in the period, while orders describe what customers have committed to buy. The two measures moving together suggests that demand is not simply working through a backlog that is about to run dry.
Profitability improved as well. The comparable operating margin came in at 11.7%. A comparable figure strips out items that management considers unrepresentative of ongoing performance, such as certain restructuring and acquisition-related charges, so it is the measure most often used when judging how the core business is performing.
3. Why does data center demand matter to a telecom vendor?
Data center demand matters because every AI cluster depends on networking gear. Thousands of processors have to exchange data at very high speed, both inside a building and between buildings, and that traffic travels over optical links and through high-capacity routers and switches. Nokia sells products of this kind, which is why the surge in AI spending reaches its income statement.
In practical terms, the networking equipment can be thought of as the plumbing that connects the servers. Chips receive most of the attention in discussions about AI, yet a cluster is only as productive as the connections that keep those chips supplied with data. As operators of large computing facilities expand capacity, they buy more of this plumbing, and the vendors that supply it see the benefit.
For Nokia, that translates into a shift in who its most important customers are. The company continues to supply communications service providers, but the customers who are increasing orders at triple-digit rates are cloud and AI companies. The company has been repositioning toward that group, and the quarter indicates the repositioning is producing measurable results.
4. How optical and IP networking fit into an AI cluster
Two product families sit at the center of this demand. Optical networking equipment moves enormous quantities of data as pulses of light over fiber, both within a campus and across long distances. IP networking equipment, meaning routers and related systems, decides where each packet of data goes and keeps traffic flowing when links become congested.
AI training and inference workloads are unusually demanding on both. A single job may be spread across a very large number of processors that must stay synchronized, so delays or dropped traffic leave expensive chips idle. Buyers therefore tend to prioritize network capacity and reliability alongside the processors themselves, which supports demand for vendors with strong portfolios in both areas.
This is the broader context for Nokia's results. The company is not selling the processors that dominate headlines about AI spending, but it is selling a layer that grows whenever those processors are deployed in greater numbers. That relationship explains why a telecom equipment maker can report triple-digit growth in a segment tied to cloud customers.
5. Guidance stays where it was
Hotard did not change the outlook for the full year. Nokia continues to expect comparable operating profit of between €2.1 billion and €2.6 billion for 2026. Keeping a range in place after a strong quarter can be read in two ways. It may reflect caution about the remaining months of the year, or it may simply mean management sees the existing range as still adequate.
Hotard pointed to continued momentum in AI and data center demand as a supporting factor for the business. Unchanged guidance means investors have no new numerical target to react to, so attention is likely to focus on whether the pace of orders is maintained in coming quarters rather than on any revision of the annual range.
The wide span between the low and high ends of the range, half a billion euros, also leaves room for the final result to vary with how quickly orders convert into recognized sales. That conversion is influenced by customer delivery schedules, which large cloud buyers set according to their own construction timelines.
6. Infinera integration ahead of schedule
Hotard also reported progress on the integration of Infinera, the optical networking company Nokia acquired. According to the company, the work is on track to be completed by the end of 2026, ahead of the originally planned timetable.
Infinera brings capabilities in optical networking and in data center interconnect, the technology that links separate data center sites over long distances. That specialty is closely tied to the AI buildout, because large computing projects frequently span several buildings that need to behave as one system.
An acquisition integration that finishes early tends to matter for two reasons. It lets the combined company present a single product set to customers sooner, and it frees management attention and engineering resources for new work. Whether the early completion will show up in future margins is something the company has yet to demonstrate in its reported numbers.
7. Reading the order figure
The 123% sales increase is the headline number, yet the €1.5 billion in orders may be the more informative one. Sales can reflect the timing of deliveries, while a large order intake indicates what customers expect to need in the quarters ahead. It suggests that the strength is not confined to a single period.
The margin result adds to that reading. A business can grow quickly while earning less on each sale, for instance when it discounts heavily to win new accounts. A comparable operating margin of 11.7%, reported alongside sales that more than doubled in the AI and cloud group, points in the other direction, indicating that the newer revenue is not arriving only as unprofitable volume.
Both readings come with the qualification that they rest on a single quarter and on figures supplied by the company. They describe a trend that is encouraging, but a trend needs several consecutive reports before it can be called established.
8. Risks: concentration and cyclical spending
The main risk is the reverse side of the opportunity. Spending on AI infrastructure is concentrated among a small number of very large companies. If a few of them were to slow their capital expenditure, scale back their data center plans or change suppliers, Nokia's AI and cloud figures would feel the effect quickly, because that segment is now responsible for a large share of its growth.
Capital spending by large technology companies is also cyclical. Periods of rapid expansion have historically been followed by pauses while buyers digest what they have installed. Orders growing at more than 100% a year are unlikely to continue indefinitely, and a more modest rate would not necessarily signal trouble, but it would change how the company is valued.
Nokia is also a business that still depends on telecom operators, whose own budgets have been under pressure. A strong data center performance can offset weakness elsewhere in the group, but it does not remove it, and results will continue to reflect the balance between the two.
9. What to watch next
Three items will indicate whether the quarter marks a durable change. The first is whether orders from AI and cloud customers keep rising after the €1.5 billion recorded in the third quarter. The second is whether Nokia lands within its 2026 comparable operating profit range of €2.1 billion to €2.6 billion. The third is whether the Infinera integration finishes by the end of 2026 as planned.
The next set of quarterly results will show how each of these developed. Until then, the third-quarter figures give a clear statement of direction: a company once defined by mobile networks is deriving its fastest growth from the infrastructure that underpins AI, and investors now have to judge how long that demand can last.