Nokia advanced its AI-RAN partnership with NVIDIA by completing GPU-accelerated RAN functional tests, integrating new operators, and staging multiple live demonstrations at Mobile World Congress 2026. The companies validated Nokia’s anyRAN software on the NVIDIA AI Aerial platform, moving AI-RAN from lab validation toward commercial deployment. Operators including BT, Elisa, NTT DOCOMO and Vodafone Group are now engaging on AI-RAN initiatives aimed at evolving 5G networks toward AI-native 6G architectures.
The collaboration centers on running AI and RAN workloads concurrently on shared accelerated computing infrastructure. Nokia confirmed successful functional tests of AI and Layer 1 RAN workloads on NVIDIA GPU-accelerated systems in real operator environments. The companies positioned AI-RAN as a shift from hardware-centric upgrade cycles to software-defined, cloud-native radio platforms capable of supporting both connectivity and AI services.
At MWC26, Nokia is showcasing AI-RAN use cases at booth 3B20 in Hall 3, including vision AI with reasoning capabilities, robotic applications, and monetization of spare GPU capacity within distributed RAN sites. The demonstrations highlight how operators can optimize energy efficiency, automate operations, and expose excess compute resources to third parties as new revenue streams.
AI-RAN technology readiness and operator milestones:
• T-Mobile U.S.: Over-the-air testing at T-Mobile’s AI-RAN Innovation Center in Seattle combined Nokia AirScale Massive MIMO radios in the 3.7 GHz (n77) C-band with concurrent AI and RAN processing on a single NVIDIA Grace Hopper 200 server. The trial supported video streaming, generative AI queries, and AI-based video captioning using commercial spectrum and devices.
• Indosat Ooredoo Hutchison: Completed Southeast Asia’s first AI-RAN powered Layer 3 5G call using an open, cloud-native network with Nokia AirScale remote radio heads and NVIDIA GPU acceleration. The test demonstrated simultaneous AI and RAN workloads on shared infrastructure in a live operator setting.
• SoftBank Corp.: Demonstrated use of spare AI-RAN compute capacity, orchestrated by SoftBank’s AITRAS Orchestrator, to run third-party AI workloads. The trial aligned with AI-RAN Alliance architectural guidance and illustrated how operators can monetize distributed RAN compute resources.
Expanding ecosystem and infrastructure stack:
• Dell Technologies, Quanta (QCT), and Supermicro are providing GPU-optimized server platforms to support AI-RAN deployments using commercial off-the-shelf infrastructure.
• Red Hat OpenShift and Red Hat AI Enterprise provide the cloud-native orchestration layer for scaling NVIDIA-accelerated RAN and AI workloads across hybrid cloud environments.
• QCT’s QuantaEdge EGN77C2U, based on the NVIDIA Aerial RAN Computer Pro platform and powered by NVIDIA RTX PRO Blackwell GPUs and the NVIDIA Grace CPU, delivers compact edge compute for inline GPU-accelerated RAN processing, integrated with Nokia’s anyRAN architecture.
Justin Hotard, President and CEO of Nokia, said: “AI is the new workload reshaping networks. That shift requires architectural change across every layer — including the radio. AI-RAN transforms RAN into a software-driven platform optimized for AI, and with NVIDIA and a growing ecosystem of partners we are progressing from validation to commercial deployment. This is a foundational step toward AI-native networks and 6G.”