T-Mobile Completes First Commercial AI-RAN Call as NVIDIA Expands Carrier Push

By AI123

On October 21, NVIDIA announced that major global telecom operators are expanding its accelerated computing technology into communications networks to deliver services closer to users and enterprises, improve wireless experiences, and automate network operations. T-Mobile has completed the world's first commercial AI-RAN call, while Deutsche Telekom, BT, and Elisa announced AI-RAN trials.

AI-RAN adds AI compute to communications networks, allowing base stations and nearby nodes to handle AI tasks locally while transmitting signals. NVIDIA CEO Jensen Huang said at the GTC conference in Berlin that telecom operators will all provide AI compute in the future, and base stations could evolve into a distributed AI network. He added that AI's next wave will enter the physical world, with communications networks becoming key infrastructure for physical AI.

Huang noted that mobile base stations themselves are an important application of physical AI. "It senses the world differently from cameras. It senses the world through radio waves, reasons, processes information, and interacts through beamforming," he said. AI will transform telecom operations and improve spectrum efficiency. The idea is not to put all AI compute into base stations, but to move latency-sensitive tasks to the network edge.

Edge computing places servers and AI inference near users and devices, reducing latency and bandwidth pressure. AI-RAN competition is still at an early stage. NVIDIA and Nokia cited an Omdia forecast that the cumulative market opportunity for AI-RAN will exceed $200 billion by 2030.

Explore topicsCompute PowerGTCNVIDIA

Sources

View all