Editorial Team
7 min read
Nokia booked €2.8 billion in AI and cloud order intake in the second quarter, as optical and IP-network sales grew. The orders strengthen its data-centre strategy, but they are not revenue yet, and restructuring, supply constraints and negative cash flow leave execution as the central question.
Nokia’s second-quarter results give its data-centre push a more tangible commercial marker: €2.8 billion of new AI-and-cloud orders across its optical and IP-networking businesses. For the Finnish network-equipment maker, that is evidence of demand beyond a one-quarter sales lift. It is not, however, the same thing as realised revenue or cash.
Justin Hotard, Nokia’s president and chief executive since April 2025, previously ran Intel’s Data Center & AI Group and led Hewlett Packard Enterprise’s high-performance-computing, AI and Labs business. That record matters because Nokia is repositioning a company better known for telecom equipment around the networks that connect AI data centres. Nokia’s leadership profile identifies those prior roles; the strategy still has to prove it can translate that shift into durable returns.
Nokia’s Q2 2026 sales to Telecommunication Providers and AI&Cloud customers, company-reported. Source: Nokia Q2 and half-year 2026 report.
Nokia said in its Q2 report that it expects around half of the €2.8 billion order intake to convert to revenue in the next 12 months. That implies roughly €1.4 billion of expected conversion, but it remains management guidance rather than booked sales. The company says customers are placing longer-term orders because supply is the industry’s main constraint.
The revenue base puts both the momentum and its limits in perspective. AI-and-cloud customers generated €446 million in Q2 sales, up 105% year on year on a constant-currency-and-portfolio basis. That is about 9% of Nokia’s €4.815 billion group sales, versus €3.514 billion, or about 73%, from telecommunications providers. The AI category is growing faster, but it has not displaced Nokia’s incumbent customer base.
Network Infrastructure is the business meant to capture that opportunity. It sells optical transport for metro, long-haul and data-centre interconnects, IP routing and data-centre switching, as well as fixed-network equipment. The segment’s Q2 sales rose 12% on a constant-currency basis to €2.037 billion; Optical Networks rose 20% to €868 million and IP Networks 16% to €679 million. Its comparable operating margin rose to 8.1% from 6.4% a year earlier.
The source of some of that change is narrower than the headline suggests. Optical Networks also benefited from telecommunications-provider sales, while IP Networks’ growth from AI and cloud was partly offset by lower telecom-provider sales. Fixed Networks declined 2% on the same currency basis. Nokia’s results presentation likewise presents the AI-and-cloud result alongside a wider portfolio plan, including divesting or separating businesses it no longer considers core.
Nokia bought scale before this quarter. It completed the acquisition of Infinera, the San Jose optical-networking company, in February 2025 and folded it into Optical Networks. In the closing announcement, Nokia said the deal broadened its webscale presence and targeted more than €200 million of net comparable-operating-profit synergies by 2027. Those are targets, not reported savings; Q2 comparable results also excluded €14 million of Infinera-related transaction and integration costs.
The company is also expanding manufacturing capacity for optical components. Nokia says its San Jose fab should begin ramping production later in the fourth quarter; advanced test-and-packaging capacity in Pennsylvania is due to increase tenfold starting in the third quarter. It has agreed, subject to approvals, to acquire NXP’s Chandler, Arizona semiconductor campus, first leasing part of it from early 2027 and expecting to convert it to indium-phosphide production; the full-site transaction is expected to close in the first quarter of 2029.
That timetable clarifies the trade-off in the order announcement. Component scarcity may encourage customers to reserve supply earlier, but it also requires Nokia to secure inputs and carry inventory. The company reported a €370 million inventory increase and a €280 million rise in receivables during the quarter. An independent analysis also noted rising memory-chip costs and long component queues across the sector; it reported that Ericsson had warned of pressure from higher component costs. The comparison is a warning about inputs, not proof that Nokia is insulated from them.
Nokia reported €434 million in comparable operating profit, up 18% year on year and above the €382 million average analyst estimate cited in a market report. But Nokia defines comparable measures by excluding, among other items, restructuring charges, acquisition-related amortisation, impairments and integration costs; the measures are not IFRS-defined and are not necessarily comparable with those of other companies.
On a reported basis, operating profit was a €50 million loss, compared with a €147 million profit a year earlier. The reconciliation includes €390 million of restructuring and associated charges, €46 million of acquisition-related amortisation and depreciation, €30 million of impairments and write-offs, and the €14 million in Infinera-related costs. Reported profit for the period was €5 million, down from €96 million.
Free cash flow was negative €732 million, versus positive €88 million a year earlier. In its cash-flow disclosure, Nokia attributes that partly to about €1.15 billion of working-capital outflows, which included receivables, inventories and lower liabilities, plus restructuring and capital spending. Net cash fell €1.012 billion from the end of March to €2.776 billion at the end of June. That does not negate the order intake, but it means a larger sales pipeline currently demands cash before it produces it.
The restructuring programme has grown too. Nokia now expects €800 million of related charges in 2026: €250 million for the conclusion of its 2023–26 programme, €350 million for integrating its China operation into the global model, and €200 million for additional actions primarily in Europe. The company targets €200 million in cost savings from the China integration, but that too is a target. The independent analysis reports that Hotard discussed possible early-retirement options in Europe and did not give a new headcount target.
Nokia’s results are not a pure data-centre story. Mobile Infrastructure was €2.680 billion of Q2 sales, larger than Network Infrastructure, and its sales grew 7% on a constant-currency basis. Yet its comparable operating margin slipped to 11.6% from 12.2%.
Within the segment, Radio Networks grew 7%, but an independent analysis reports that Hotard attributed part of the year-on-year comparison to earlier-than-usual recognition of RAN software revenue. The analysis also says Nokia’s performance came as the broader RAN market had weakened and Ericsson’s comparable radio sales had declined. That makes it premature to treat one quarter’s radio growth as evidence of a broad market recovery.
Nokia’s full-year comparable-operating-profit outlook is €2.1 billion to €2.6 billion, but the company says the €100 million increase from its prior range is technical: two businesses were reclassified as discontinued operations. Operationally, the outlook is unchanged. It assumes 3% to 7% sequential sales growth in Q3, broadly flat comparable operating profit because of software-revenue timing, and a meaningful increase in Q4.
The next evidence should be concrete: whether the expected roughly €1.4 billion of AI-and-cloud order intake becomes revenue on the stated timetable, whether the manufacturing build-out reduces rather than deepens supply exposure, and whether receivables and inventory ease without derailing free-cash-flow conversion.
Until then, the quarter supports a narrower conclusion than an AI-led turnaround. Nokia has a growing optical and IP order pipeline and a strategy shaped by a chief executive with data-centre experience. It also has a customer mix still dominated by telecom providers, a radio business affected by revenue timing, and a transformation whose restructuring and working-capital costs remain visible in reported earnings and cash flow.
AMD has introduced ROCm.AI to guide setup, deployment and optimization on its hardware, while HIP and HIPIFY offer a separate source-level route from CUDA for embedded software. The unresolved question is whether those tools lower enough engineering risk to matter for large planned AI deployments.
KTransformers v0.6.4 adds LoRA and hybrid fine-tuning to its CPU-GPU workflow for sparse mixture-of-experts models, expands an Ampere inference path and moves a scheduler endpoint to loopback after a report of unauthenticated pickle deserialization. Its published throughput figures are configuration-specific release claims, not a general comparison of fine-tuning methods.
Microsoft has put MAI-Image-2.5-Pro and MAI-Voice-2-Flash into public preview in Foundry. The launch extends the company’s in-house AI options, but its evidence for lower cost and broader deployment is chiefly tied to MAI-Image-2.5 and company-reported workloads—not an independently comparable test of the new image model.
Microsoft has made its MAI-Code-1-Flash coding model available in paid GitHub Copilot plans and says a related model is live in Excel. Its evidence points to a targeted effort to reduce serving costs, while leaving the scale, task coverage and economics of the broader rollout unresolved.
AMD has detailed the 256-core EPYC 9996, its density-oriented Venice server CPU, alongside a broader 9006 platform with 16 memory channels and PCIe 6.0. But the Helios AI rack uses a separate high-frequency Venice design, and AMD’s largest rack-performance figures remain preliminary company projections rather than independent results.
NVIDIA says it will commit $4 million over three years to let CNCF projects test on real GPUs. With Kubernetes’ core Dynamic Resource Allocation framework already generally available, the consequential question is whether shared validation can make diverse GPU implementations dependable without settling operators’ choices over isolation and control.
Microsoft has made MAI-Image-2.5 the end-to-end default for Bing Image Creator and put it into PowerPoint image-to-image features. The company says the PowerPoint deployment can reduce GPU costs by up to 84% versus GPT-Image-2, while its much larger OpenAI partnership remains in place.
Centrica says its roughly 1,300 proposed role reductions reflect lower customer contact and a wider transformation programme. The record shows AI is one tool in that programme, while the headline total reaches beyond the 500 contact-based roles it has specifically identified.
Meta is using a new mass-market film to argue that AI will extend its mission of connection. The company has real distribution, product and infrastructure plans behind the pitch, but the meaningful test is whether it makes access, controls and outcomes legible as those plans reach users.
Google is rolling out an optional selfie-video check for eligible Google Account holders who lose access to their usual devices. The feature adds a biometric reference to recovery, but it excludes several account types and leaves users dependent on Google’s unreported real-world performance and recovery decisions.
Google’s first ATLAS report finds Gemini use across occupations covering 88.4% of U.S. employment, but mostly in a limited share of each job’s tasks. The data offer a detailed view of how Google’s AI is used; they do not show whether it is raising output, eliminating work or changing hiring.
Super Micro Computer says it received more than $60 billion in fiscal fourth-quarter orders and now expects 15% to 17% gross margins, despite revenue tracking near the low end of its outlook. The preliminary figures point to a much larger AI-server delivery pipeline, but not yet to booked revenue or a durable improvement in profitability.
Nebius says it has brought up its first full NVIDIA Vera Rubin NVL72 rack in Finland. The deployment is an early integration milestone backed by a $2 billion NVIDIA private placement, but the company still has software and production testing to complete before customers can use it.
According to people familiar with the effort, Amazon founder and executive chairman Jeff Bezos has pressed Prime Video to make AI and personalization central to a proposed redesign known internally as Lighthouse. The early test could improve discovery, but Amazon has not said how it would reconcile personalized rankings with the paid placement and subscriptions that make Prime Video a complex storefront.
A report says some Chinese cities are gradually issuing robotaxi permits again after a review prompted by Baidu’s Wuhan outage. Shenzhen’s revised rules show that road testing, demonstrations and unmanned trials remain separately governed—and do not by themselves establish broad commercial service.
Databricks says it will run core operations and analytics on Azure Databricks, deepen Microsoft product integrations and expand use of Azure Cobalt chips. The partnership runs into the 2030s, but the companies have not disclosed its value, workload scale or customer results.
OpenAI is beginning a U.S. rollout of Health in ChatGPT for logged-in adults on web and iOS. The feature can bring optional Apple Health and supported medical-record context into ordinary chats, but connected data may be incomplete and requires user permission before use.
Yelp will supply ChatGPT with reviews, ratings, photos and business details, with branded links back to Yelp and a planned quote-request feature. The non-exclusive deal extends an existing licensing strategy, but neither company has disclosed its financial terms or how ChatGPT referrals will be measured and sold.
U.S. stocks fell as attacks on Saudi tankers added a second threat to Middle East oil flows and earnings from Alphabet and Tesla highlighted the cash demands of the AI buildout. The selloff was concentrated in megacap technology rather than a uniform retreat from equities.
Blackstone’s second-quarter distributable earnings rose 26% as realizations and fee-related earnings increased. Its data-center and other AI-linked investments are becoming a larger part of the firm’s portfolio narrative, but the results do not separately show how much of the gain came from AI while retail private-credit fundraising slowed.