July 2026 - Dell'Oro Group

Nokia’s latest AI-RAN announcement is more about increased confidence and improved messaging than a fundamental change in strategy. The AI-RAN roadmap and timeline remain largely unchanged, but Nokia has refined its migration story, sharpened its focus on software-driven innovation, and provided additional insight into how it views AI-RAN’s role in addressing operator challenges and strengthening its competitive position in the evolving RAN market.

Nokia’s AI-RAN roadmap and timeline remain broadly unchanged. The AI-RAN roadmap and timeline are largely unchanged from what the company shared with analysts and investors over the past year, suggesting Nokia is executing against an existing strategy rather than introducing a new one.

The migration path has been refined and simplified. Compared with last year’s Nokia-NVIDIA framework—which highlighted purpose-built D-RAN, D-RAN vRAN, and C-RAN vRAN—the company has simplified the way it presents deployment options. Nokia is now placing less emphasis on specific vRAN architectures and more emphasis on deployment flexibility and performance profiles (installed-base leverage, high-capacity AI-RAN, and cloud-native AI-RAN). Broadly speaking, the migration strategy remains intact, with the changes reflecting refinements in packaging and messaging rather than a fundamental shift in direction.

Nokia is becoming more confident in its merchant silicon/GPU-based AI-RAN strategy. The most notable changes are in messaging and conviction. There was more hedging at MWC Barcelona when analysts pressed them on their 6G R&D efforts between custom and merchant silicon. Now, Nokia appears increasingly confident in its merchant silicon/GPU-based approach and is positioning AI-RAN as a software-defined platform strategy rather than a hardware story. While the company continues to support multiple hardware options, management indicated that most future software innovation and feature development will target the merchant silicon track.

AI-on-RAN is also becoming more tangible. Beyond AI-for-RAN use cases, Nokia is now highlighting sensing, positioning/location services, and third-party software applications that could run on the platform. Importantly, these are being positioned as complementary software-ecosystem opportunities rather than as the primary justification for the architecture.

The 2x spectral efficiency claim is compelling but requires more context.The headline claim of up to 2x spectral efficiency improvements is compelling, especially since the gains appear to exceed those highlighted in other AI-RAN announcements. At the same time, it should be viewed with some caution until there is greater clarity around the underlying assumptions, benchmarks, and how it compares with competing approaches from Huawei, Ericsson, and others in actual deployments. Spectral efficiency remains highly dependent on deployment scenarios and starting points, making cross-vendor comparisons difficult. Hopefully, Ookla can update this RAN efficiency chart by 2028.

The announcement supports Dell’Oro’s AI-RAN and GPU-RAN forecasts. More broadly, the announcement is consistent with our recently published $35 B AI-RAN forecast and our upward revision to the GPU-RAN outlook, which now exceeds $1 B by 2030. Nokia’s emphasis on AI-for-RAN as the primary value proposition, combined with its growing confidence in merchant silicon-based deployments, is consistent with our view that AI-for-RAN adoption is accelerating as operators prepare for more software-centric and AI-native RAN architectures.

AI-RAN is strategically important to Nokia’s broader RAN turnaround efforts. Finally, Nokia’s AI-RAN strategy should be viewed in the context of its broader RAN turnaround efforts. Having lost approximately 10 percentage points of RAN market share over the past decade, Nokia is well aware of the importance of maintaining sufficient scale in the highly concentrated RAN market. AI RAN is therefore more than a technology roadmap—it represents a strategic bet that software-driven innovation, flexible deployment options and AI-native capabilities can help strengthen Nokia’s competitive position over the long term. The latest announcement signals that Nokia is becoming increasingly confident that AI-native RAN will play a central role in its efforts to improve its competitive standing in the RAN market.

Ericsson’s Mobility Report provides a comprehensive overview of the end-user trends shaping the RAN market. The report is loaded with data and insights. Below are six takeaways we are keeping in mind at Dell’Oro Group as we update our 5-year RAN forecast.

22% mobile traffic growth in 1Q26

Mobile traffic growth continues to moderate. Given concerns that traffic growth could decelerate more quickly than expected, it is encouraging that total mobile network traffic has now posted seven consecutive quarters of stable year-over-year growth in the 20% to 22% range. While growth rates are no longer at the elevated levels seen in the early days of 5G, traffic demand is growing and continues to support long-term RAN capacity investments.

Total traffic to grow 2.5x by 2030

Ericsson projects total mobile network traffic (mobile broadband and FWA) to increase by approximately 2.5x between 2025 and 2030, implying roughly 20% annual growth. This outlook influences not only the timing of 6G deployments but also the level of capacity investments during the second half of the 5G cycle. Even as the industry is trying to pivot towards network differentiation and the Uplink (UL), the relationship with RAN revenue and the overall mobile traffic growth is expected to improve, especially as the industry is now in somewhat uncharted territories when it comes to the supply/demand of mobile connectivity.

UL 10% to 12% of total traffic

For years, suppliers have emphasized the importance of improving UL performance. First, it was short-form video, and now AI-driven applications are expected to further increase UL demand. Even so, our interpretation of Ericsson’s medium- and high-UL scenarios is that UL traffic will still account for only about 10% to 12% of total mobile network traffic by 2031, up from roughly 8% today. AI may significantly reshape traffic patterns, but Downlink (DL) traffic will continue to dominate overall network demand.

10 M smart glasses in 2025

The projected growth in mobile traffic continues to be overwhelmingly driven by smartphones. Although expectations are rising that smart glasses could become the next major connected device category, Ericsson estimates shipments reached approximately 10 million units in 2025—roughly 1% of smartphone shipments. The figure is an important reminder that while the long-term opportunity is compelling, adoption remains in its early stages, and the base case is that smartphones will continue to dominate mobile traffic for the foreseeable future.

Satellite accounts for 2% of fixed connections

Even as Starlink and other satellite providers continue to expand their ambitions across home broadband, enterprise networking, and mobile connectivity, Ericsson projects satellite broadband will account for about 2% of global fixed broadband connections by 2031. In comparison, Fixed Wireless Access (FWA) is expected to represent approximately 17% of fixed broadband connections, reinforcing that terrestrial wireless technologies remain the much larger opportunity.

80% of enterprises are growing their mobile investments

Private wireless increased 16% and accounted for 3 to 5% of total RAN revenue in 2025. To date, the primary driver has been industrial 4G and 5G deployments supporting applications where Wi-Fi or public cellular networks cannot meet performance, coverage, or mobility requirements. Looking ahead, Ericsson’s survey of more than 100 enterprise decision makers across North America, Europe, and Asia suggests that 80% plan to increase their mobile investments, in part to help scale AI initiatives. While private wireless remains a relatively small portion of the RAN market today, enterprise demand continues to strengthen.

In other words, AI continues to dominate industry discussions; however, the latest Ericsson Mobility Report reinforces that the RAN market over the next five years will still be driven primarily by steady traffic growth, smartphones, FWA, and enterprise investment. AI is expected to increase UL requirements and create new network demands, but the fundamental traffic drivers remain largely unchanged.

Contact Us