NVIDIA Moves AI-RAN to Field Deployments With Global Partners at MWC 2026

NVIDIA and global telecom giants announced a major leap for AI-RAN technology at Mobile World Congress 2026. The transition from laboratory testing to live field environments proves that software-defined architectures are ready for prime time. Partners like T-Mobile U.S., SoftBank, and Nokia are leading this charge across Europe, Asia, and North America. Technical benchmarks showcased at the event reached unprecedented levels of performance. Partner SynaXG achieved 36 Gbps throughput with sub-10 millisecond latency on a single server. This marks the world first implementation of AI-RAN on millimeter wave bands running alongside sub-6GHz spectrum.

In Southeast Asia, Indosat Ooredoo Hutchison completed the region's first AI-powered 5G call. This milestone included the real-time remote control of a robotic dog over a live network. The demonstration highlights how network intelligence can manage complex, low-latency tasks in real-world conditions.

SoftBank reached another milestone with its AITRAS platform by achieving a 16-layer massive MIMO configuration. This breakthrough proves that spectral efficiency can be significantly improved through software-defined 5G on general-purpose GPUs. Operators are no longer restricted by specialized hardware for high-capacity urban environments.

The ecosystem is expanding rapidly with new hardware from Quanta Cloud Technology and Supermicro. These companies are providing standardized, off-the-shelf servers for NVIDIA AI Aerial platforms. This standardization allows operators to choose from various deployment solutions while maintaining a common software foundation. NVIDIA is also accelerating the path toward 6G by joining the OCUDU Ecosystem Foundation. By contributing open-source libraries, the company aims to foster a collaborative research environment. This move is supported by industry sentiment, as 77 percent of telecom leaders expect a faster 6G deployment cycle.