Ericsson Demonstrates Cloud RAN Software on NVIDIA AI Infrastructure with T-Mobile

Ericsson ’s RAN software strategy enables mobile operators to run the same radio access network (RAN) software on both Commercial Off the Shelf (COTS) hardware and purpose-built hardware using Ericsson Silicon. When running on COTS hardware, the software can utilize Selected Function Accelerator hardware and software – when available – for compute-intensive tasks.

In collaboration with T-Mobile,  Ericsson has successfully demonstrated its Cloud RAN software on NVIDIA  AI infrastructure using the NVIDIA Aerial CUDA. This shows Ericsson’s commitment to providing mobile operators with a hardware-agnostic RAN stack that now also supports NVIDIA’s accelerated computing platform – offering maximum flexibility.

“Cloud RAN software is portable by design. By running the same RAN software stack across multiple hardware platforms, we reinforce our commitment to providing mobile operators with true flexibility without compromising on high performance.”

– Mårten Lerner, Head of Networks Strategy & Product Management, Business Area Networks, Ericsson

“At T-Mobile, every network decision starts with customer centricity, ensuring our investments deliver the best performance, reliability, and value for those who trust us with their connectivity. Our over-the-air trial in the AI-RAN Innovation Center in Bellevue demonstrates how we are evolving from a connectivity pipe to an intelligent platform, laying the groundwork for AI-native services and future innovation in advanced AI and 6G. Portability across compute platforms strengthens our flexibility and resilience, allowing us to deploy the right technology for each use case and scale with disciplined execution behind our unmatched combination of the Best Network, Best Value, and Best Customer Experience.”

– Ankur Kapoor, Chief Network Officer, T-Mobile

An open ecosystem, built for now and next

• One software stack, multiple options: Ericsson Cloud RAN runs across Ericsson Silicon, NVIDIA AI Infrastructure, and other hardware platforms, giving operators the potential for control and commercial flexibility.
• Selected Function HW acceleration for efficiency: Offloading compute-intensive functions and AI-for-RAN functions to NVIDIA AI Infrastructure
• Portability by principle: Mobile operators can deploy the right compute at the right site, at the right time, across centralized and distributed sites.

Ericsson will showcase the Cloud RAN demo on NVIDIA AI Infrastructure at Mobile World Congress 2026 in Barcelona, taking place from March 2 to 5.