The Pulse
Nokia reports 20% spectral-efficiency gain as AI-RAN trials spread
Nokia says its AI-RAN platform has improved spectral efficiency by more than 20% as operators across four regions move into proofs of concept and live trials. The company is pairing its anyRAN software with NVIDIA Aerial RAN Computer for 5G
AI.info Team ·
Nokia says its AI-RAN platform has delivered more than 20% higher spectral efficiency as operators across North America, Europe, Asia-Pacific and the Middle East move from early evaluations into laboratory and live field trials.
The September 16 announcement names A1 Group, Chunghwa Telecom, du, e&, Mobily, stc, TPG Telecom and Zain Saudi among operators advancing proofs of concept with Nokia and NVIDIA Aerial RAN Computer. Nokia says earlier trials with NTT DOCOMO continue, while evaluations involving T-Mobile, SoftBank and Indosat Ooredoo Hutchison add to the company’s existing operator base.
The figures and deployment claims come from Nokia’s own announcement. The company presents the activity as evidence that AI-RAN is moving toward commercial deployment, but the release does not provide independent test results, deployment volumes or revenue figures from the named operators.
More than 20% efficiency from software
Nokia says its AI-RAN platform combines its AI-native anyRAN software with NVIDIA Aerial RAN Computer. The arrangement is intended to make radio access networks more programmable, allowing the same infrastructure to handle conventional connectivity alongside AI workloads.
Higher spectral efficiency would let operators carry more traffic through existing spectrum and network infrastructure. Nokia says the approach could improve capacity and network economics while reducing dependence on traditional hardware refresh cycles, although the announcement does not specify the test conditions behind the reported improvement.
The company also says its software roadmap targets additional efficiency gains in 2027 and 2028. No numerical targets for those future gains appear in the release.
Eight operators join the trial list
The newly named operators cover a wide geographic range and include mobile groups in Europe, Taiwan, the Gulf, Saudi Arabia and Australia. Nokia describes their work as a mix of proofs of concept, lab work and live field trials rather than a single coordinated commercial rollout.
That distinction matters. A proof of concept can test a narrow radio or software function, while a live field trial places the system under operational network conditions. Nokia does not disclose which named operator is conducting which type of trial, nor does it give launch dates for commercial services.
Previous work with T-Mobile, SoftBank and Indosat Ooredoo Hutchison forms part of Nokia’s broader effort to establish an operator and equipment ecosystem around AI-RAN. The company also points to a growing group of original equipment manufacturers and expanded cooperation with NVIDIA on platform and software development tools.
Emma Falck frames radio networks as programmable systems
AI is reshaping the future of telecommunications and creating a significant opportunity for operators to rethink the role of the radio network. Together with NVIDIA and our growing ecosystem of partners, we are helping operators transform their networks into intelligent, programmable platforms that can support both connectivity and AI services.
The increasing engagement we are seeing from service providers around the world shows that AI-RAN is moving quickly from vision to commercial reality and becoming a foundational technology for the AI-native 6G era.
Emma Falck, President of Mobile Infrastructure at Nokia
Falck’s statement places AI-RAN inside Nokia’s wider transition from fixed-purpose radio equipment toward software-defined systems. The company says its architecture supports 5G today and provides a software upgrade path toward 6G, though standards, hardware requirements and commercial timelines for 6G remain unsettled.
NVIDIA wants the radio network to carry AI workloads
NVIDIA Senior Vice President of Telecom Ronnie Vasishta describes the combination as a way to turn mobile networks into distributed AI infrastructure. The company’s Aerial RAN Computer supplies the computing platform, while Nokia provides the radio software and network expertise.
Nokia says rising AI traffic is changing network requirements through heavier uplink demand, new traffic patterns and a need for computing resources closer to users. Those pressures give operators a reason to seek capacity gains without relying only on new spectrum or large equipment replacements.
Omdia Practice Leader for Mobile Infrastructure Remy Pascal says the breadth of Nokia’s operator engagement indicates that AI-RAN is becoming a strategic priority rather than remaining limited to research projects. He also says cooperation among operators, vendors and other participants will determine how much of the technology’s potential reaches production networks.
The unanswered commercial questions
Nokia’s announcement establishes the scale of its current operator engagement and its claimed efficiency result, but it leaves several commercial questions open. The release does not identify contract values, production deployment counts, customer spending or the specific network metrics behind the 20% improvement.
It also does not say when any of the newly named trials will become commercial services. For now, the clearest milestone is the shift from laboratory evaluation toward field testing across multiple regions, supported by Nokia software and NVIDIA computing hardware.