AI-native networks have long been positioned as a 6G ambition. At MWC 2026 in Barcelona, that vision shifted into operational reality. Major telecom vendors, chipmakers, and operators presented field trials, commercial launches, open-source toolkits, and multi-operator coalitions focused on AI-RAN and AI-native 6G infrastructure.
The most consequential move came from Nvidia, which secured commitments from BT Group, Deutsche Telekom, Ericsson, Nokia, SK Telecom, SoftBank, T-Mobile, Cisco, and Booz Allen to build 6G on open, secure, AI-native software-defined platforms. Nvidia expanded its AI-RAN Alliance beyond 130 participants and released telecom-focused tools, including a 30B-parameter Nemotron Large Telco Model and blueprints for RAN energy efficiency and network configuration. Early deployments are already underway in Africa and Japan.
Nokia and Nvidia validated AI-RAN outside laboratory settings. At T-Mobile’s Seattle innovation center, Nokia’s AirScale Massive MIMO operated concurrent AI and 5G workloads on a single Nvidia Grace Hopper server. Indosat Ooredoo Hutchison achieved Southeast Asia’s first AI-RAN-powered Layer 3 5G call with shared GPU infrastructure. SoftBank demonstrated monetizing spare RAN compute capacity via third-party AI workloads.
Ericsson introduced a competing model. Instead of GPUs, Ericsson embedded neural accelerators directly into purpose-built silicon. Its AI-ready radios deliver AI-managed beamforming and latency-optimized scheduling while emphasizing total cost efficiency. Ericsson simultaneously expanded collaboration with Intel to accelerate AI-native 6G readiness.
Beyond telecom infrastructure, AI-native networks reshape enterprise connectivity. GPU-embedded RAN architecture enables AI inference at the edge, closer to data sources. Networks increasingly resemble cloud platforms—software-defined, continuously evolving, and monetizable beyond connectivity.
MWC 2026 confirmed that AI-native networks are no longer conceptual. Live trials, commercial hardware, and operator commitments demonstrate that AI-RAN is entering production. The remaining questions concern speed of deployment, architectural models, and which ecosystem strategy will define 6G leadership.
Quelle:
https://www.artificialintelligence-news.com/news/ai-native-networks-mwc-2026/
