This Former Smartphone Maker Has Quietly Become a Top AI Stock

Nokia’s AI-RAN is positioned to unlock real-time AI inference by reducing data-center latency, with the company targeting customer trials later this year and 10 customers publicly committed. Nokia reported only 4% YoY revenue growth in Q1, but its small AI/cloud segment grew 49% and margins remain below 3%, while Nvidia will invest $1B in Nokia to support AI-native 5G-advanced and 6G upgrades. Dell’Oro estimates cumulative AI-RAN spending could reach $35B over five years, and the article notes investor accumulation ahead of commercialization, implying upside optionality even though earnings have not yet reflected the opportunity.

Analysis

This is less a near-term revenue story than a standards-and-ecosystem optionality trade. If edge inference shifts from pilot language to budgeted carrier capex, telecom gear stops trading like a low-growth utility and starts trading like a platform enabler with a higher terminal multiple. The catch is timing: trials can surface quickly, but enterprise and carrier procurement cycles usually lag by 2-4 quarters, so the market may be ahead of realized orders.

The cleaner beneficiary may be NVDA, not NOK. If micro-data centers proliferate at the tower edge, the value accrues to the accelerator/software stack and systems integration layer, while Nokia captures only a slice of the equipment and software economics. Ericsson is the most obvious competitive casualty if Nokia becomes the default AI-RAN reference, but the bigger second-order effect is on carriers and tower-adjacent infrastructure: added power, cooling, and backhaul spend could benefit AMT/CCI and fiber providers if the thesis moves from lab to rollout.

Contrarian view: consensus may be overestimating how quickly latency pain translates into revenue. The first leg is narrative, the second is trial conversion, and the third is real spend; the market is currently pricing all three as if they are one event. Falsifiers: no meaningful order commentary by the next 2 earnings cycles, no margin lift above ~3-4%, or a competitor announcement that dilutes Nokia’s first-mover advantage. In that case, the stock has more re-rating risk than operational downside protection.

Key Decisions for Investors

• Start with only a small tactical long NOK via a 6-12 month call spread, not outright size; this is a commercialization option, not a confirmed earnings inflection. Add only if trial-to-order conversion shows up in guidance or backlog.
• Pair trade: long NOK / short ERIC for 1-3 months into AI-RAN headlines. Use it as a relative-mindshare trade; cut if Ericsson announces comparable customer trials or Nokia’s pipeline commentary stays vague.
• Maintain/accumulate NVDA on pullbacks as the higher-quality way to own edge inference monetization over 6-12 months; NVDA captures more of the hardware/software stack than Nokia does.
• Set a watch item on AMT/CCI and fiber proxies rather than a trade today. If carrier budgets begin to include edge power/backhaul spend, those names become the second-order winners.