Nokia has announced major progress in its strategic AI-RAN partnership with NVIDIA, unveiling new operator integrations, successful GPU-accelerated AI-RAN functional tests, and advanced demonstrations at Mobile World Congress 2026. The collaboration highlights how AI-driven radio access networks are reshaping 5G infrastructure and laying the technological groundwork for AI-native 6G.
Building on their partnership established last year, Nokia and NVIDIA are working with leading global operators such as BT Group, Elisa, NTT DOCOMO, and Vodafone Group to deploy AI-RAN technologies. These deployments are designed to enhance network efficiency, improve spectrum utilization, and support the rapid growth of mobile AI traffic.
GPU-Accelerated AI-RAN Validated for Commercial Readiness
Nokia confirmed the successful completion of functional tests combining AI and RAN workloads on NVIDIA’s GPU-accelerated AI-RAN platform. The validation demonstrates that AI-powered RAN solutions can operate efficiently at scale, delivering enhanced user experiences and unlocking new monetization models through distributed AI services.
At T-Mobile U.S.’ AI-RAN Innovation Center in Seattle, Nokia and NVIDIA conducted over-the-air tests using Nokia’s AirScale Massive MIMO radio in the 3.7 GHz n77 band. The trial demonstrated simultaneous AI and RAN processing on a single NVIDIA Grace Hopper 200 server. Use cases included video streaming, generative AI queries, and AI-driven video captioning, proving that advanced radio functions and AI applications can coexist on shared accelerated computing infrastructure.
In Southeast Asia, Indosat Ooredoo Hutchison achieved the region’s first AI-RAN powered Layer 3 5G call at MWC. Using Nokia’s AirScale radio heads and GPU-accelerated RAN software, the milestone validated simultaneous AI and RAN workloads on shared GPU infrastructure in a live operator environment.
Meanwhile, SoftBank Corp. demonstrated how spare AI-RAN compute capacity, managed by its AITRAS Orchestrator, can run third-party AI tasks. This approach transforms RAN into an AI-enabled service platform capable of generating revenue beyond traditional connectivity.
Expanding the AI-RAN Ecosystem
Nokia is broadening its AI-RAN ecosystem with infrastructure partners including Quanta Cloud Technology and Supermicro, alongside Dell Technologies. The ecosystem is built on a cloud-native orchestration layer powered by Red Hat OpenShift, enabling operators to deploy AI and RAN workloads across hybrid cloud environments.
By integrating NVIDIA-accelerated computing with Nokia’s anyRAN architecture, operators gain flexible, COTS-based deployment options spanning Cloud RAN and purpose-built RAN environments. This architecture supports a smooth transition from advanced 5G deployments toward AI-native 6G networks.
AI-RAN Demonstrations at MWC26
At MWC26 in Barcelona, Nokia is showcasing live AI-RAN demonstrations at its Hall 3 booth. The demonstrations highlight cognitive, software-defined radio systems capable of supporting Vision AI applications, autonomous robotics use cases, and network sensing capabilities that transform radio infrastructure into a distributed intelligence platform.
Nokia is also demonstrating how unused GPU capacity within distributed AI-RAN environments can be monetized by offering AI compute services to external enterprises. These capabilities underscore AI-RAN’s potential to optimize network operations, improve energy efficiency, and enable new AI-driven services.
Industry Momentum Toward AI-Native Networks
Industry analysts note that the growing participation of American, Asian, and European operators signals that AI-RAN is emerging as a strategic industry direction. GPU-accelerated RAN testing across multiple configurations marks a significant maturity milestone, transitioning AI-RAN from validation phases toward commercial deployment.
With AI increasingly becoming the dominant workload across digital infrastructure, the Nokia and NVIDIA partnership reflects a broader shift toward software-defined, cloud-native, and AI-integrated radio networks. As operators modernize their architectures, AI-RAN is positioned as a foundational technology for autonomous networks, distributed intelligence, and the evolution to AI-native 6G.
