Booz Allen, BT Group, Cisco, Deutsche Telekom, Ericsson, MITRE, Nokia, ODC, SK Telecom, SoftBank and T-Mobile sign up to build 6G infra on open, secure AI-native platforms
The leader of the world’s biggest company, NVIDIA, wants to build a new telecoms architecture for 6G which will be AI-native and software defined. Booz Allen, BT Group, Cisco, Deutsche Telekom, Ericsson, MITRE, Nokia, OCUDU Ecosystem Foundation, ODC, SK Telecom, SoftBank Corp. and T-Mobile have signed up.
Or as Jensen Huang, founder and CEO of NVIDIA (pictured) put it, “AI is redefining computing and driving the largest infrastructure buildout in human history — and telecommunications is next”.
The thinking is that beyond old connectivity, 6G wireless networks will become the fabric for physical AI – the next iteration of AI after agentic. NVIDIA defines it as “enabling autonomous systems like cameras, robots, and self-driving cars [to] perceive, understand, reason, and perform or orchestrate complex actions in the physical world”.
If 6G networks are built on AI-RAN architecture, they will continuously evolve through software, enabling real-time intelligence and rapid development. This invites an ecosystem of participants to contribute through open and programmable platforms. Participants could include big operator groups, technology providers startups, researchers and developers.
“With a global coalition of industry leaders, NVIDIA is building AI-RAN to transform the world’s telecom networks into AI infrastructure everywhere.” Huang said. Which means embedding AI in the RAN, but also the edge and core of 6G networks.
Empirical evidence
The announcement was not only a statement of intent, it had empirical backing in the shape of a slew of demos and field trials. In a statement Nokia said, “Building on the strategic partnership announced last year, Nokia and NVIDIA are working with leading operators, including BT, Elisa, NTT DOCOMO and Vodafone Group to adopt AI-RAN technologies that enhance network performance and support the explosive growth in mobile AI traffic”.
For example…
T-Mobile successfully tested GPU-accelerated AI-RAN workloads at its AI-RAN Innovation Centre at its Seattle HQ. In T-Mobile’s over-the-air (OTA) lab environment, Nokia’s AirScale Massive MIMO radio operating in the 3.7 GHz (n77) band supported user equipment (UEs) running use cases such as video streaming, generative AI queries, and AI-based video captioning. The trial demonstrated concurrent AI and RAN processing on a single NVIDIA Grace Hopper 200 server using accelerated AI-RAN workloads, highlighting the ability to combine advanced radio access network functions with AI applications on a shared accelerated computing platform.
Indosat Ooredoo Hutchison (IOH) achieved Southeast Asia’s first AI RAN–powered Layer 3 5G call at MWC. The call used IOH’s open, cloud native network with Nokia’s AirScale remote radio heads (RRHs) and RAN software accelerated by NVIDIA GPUs. Nokia says this proves that AI and RAN workloads can run simultaneously on shared GPU infrastructure in a live operator environment, paving the way for distributed AI intelligence that makes 5G networks more efficient and intelligent and sustainable.
SoftBank Corp. has demonstrated 16-layer Multi-User MIMO (MU-MIMO) downlink in outdoor, real-world conditions with NVIDIA’s Grace Hopper 200 chip-based hardware running Layer 1 and AI workloads at the same time. The trial used a fully software-defined 5G virtualised RAN (vRAN) that was about three times more spectrally efficient and three times faster throughput than conventional 4-layer configurations.
SoftBank is using AI to analyse user traffic data to dynamically allocate communication resources (time, frequency, and spatial dimensions) in real-time. This work is part of a broader industry push toward AI-native 6G and next-generation 5G, with partners including Nokia.
