Mobile operators are experiencing a surge in AI activity on handsets and in business systems, placing new demands on the network edge and prompting operators to move processing closer to cell towers and antennas. At the same time, they must balance staff skill development, data governance, and aging infrastructure.
NVIDIA and Nokia have announced a strategic partnership to run AI workloads within radio access networks, and NVIDIA plans a potential $1 billion investment in Nokia, subject to regulatory approval. The companies say this collaboration could help operators evolve from current 5G deployments to 5G-Advanced and, eventually, 6G by bringing compute resources nearer to network nodes.
The market for AI-enabled radio networks is expanding rapidly. Analyst firm Omdia projects the opportunity could exceed $200 billion by 2030 as operators look to edge computing to boost performance and enable new services.
T-Mobile US intends to trial this approach on its network from 2026, assessing how on-network AI can support emerging features such as augmented reality and connected vehicle services. Running AI near the radio can help maintain consistent quality during traffic spikes and enable more automated capabilities in industrial and urban systems.
Consumer trends support this shift: roughly half of ChatGPT’s 800 million weekly active users access the service on mobile devices, and more than 40 million people download the mobile app each month. As more AI agents run on phones, latency and responsiveness become shared responsibilities for application developers and telecom operators alike.
NVIDIA is positioning its new Aerial RAN Computer Pro (ARC-Pro) as a central component of this transformation. The ARC-Pro integrates compute and radio functions on a single platform and is designed to be upgraded through software updates rather than frequent hardware refreshes. Nokia plans to integrate ARC-Pro capability into its radio portfolio, and Dell Technologies will support deployments with PowerEdge servers. This reflects a broader industry trend in which cloud providers, telecom vendors, and data centre suppliers adapt to handle AI traffic both in the cloud and at the network edge.
Legacy infrastructure presents clear challenges when new edge compute elements must interoperate with older equipment. Radio engineering teams will increasingly manage AI workloads, requiring new skills and operational practices. Local processing may handle private or regulated data, increasing the importance of robust data governance, privacy controls, and compliance procedures.
Nokia says its anyRAN software can reduce the cost and complexity of this transition by enabling operators to combine cloud-based and purpose-built systems. A modular design lets operators retain existing cards while adding new functions over time, limiting disruptive, large-scale replacements.
Beyond node-level compute, the companies will collaborate on data centre switching, telemetry, and optical components. Telecom operators are adopting an edge cloud model that demands seamless interoperability with cloud platforms and AI services from major providers.
For enterprise customers, advanced edge networks can accelerate analytics in logistics, improve the reliability of remote video collaboration, and enable privacy-sensitive processing for regulated industries. They will also support services for connected vehicles and industrial automation. Realizing these benefits depends on careful planning, vendor selection, and strong operational oversight.
People remain central to the shift. Operators must retrain staff for software-defined systems, procurement teams need to update vendor evaluation criteria, and finance leaders must balance potential new revenue streams against energy consumption and capital expenditure.
What leaders should think about
- Assess where edge computing can reduce latency or relieve peak traffic on the network.
- Review data governance and controls for AI workloads that operate outside core data centres.
- Favor systems that scale through software updates rather than frequent, large hardware upgrades.
- Create training programs that combine radio engineering knowledge with AI operations and data management skills.
- Develop commercial offerings for enterprise customers that require low-latency processing at the network edge.
Research into 6G is accelerating, and early work emphasizes closer integration between radio and compute. Operators that begin preparing now may unlock new revenue opportunities at the network edge as AI adoption grows. Those who delay risk remaining limited to basic connectivity while the market evolves around them.
(Image by Nvidia)
Want to discover how IoT is transforming telecoms and connectivity? Attend the IoT Tech Expo in Amsterdam, California, or London to explore how developments in 5G, edge computing, and IoT are shaping future networks and services. The event is part of TechEx and is co-located with other leading technology conferences.
Telecoms News is produced by TechForge Media. Explore other upcoming enterprise technology events and webinars through TechForge Media’s events listings.