Nokia Adopted Nvidia GPUs for AI-RAN Development
The telecommunications firm integrated the E3 interface to support artificial intelligence applications in 5G networks.
Updated on Sept. 30, 2026 in Artificial Intelligence
Live Poll
Is now a good time for mobile providers to prioritize faster network tech over broad compatibility?
Nokia has adopted the E3 interface and Nvidia GPUs to facilitate AI-driven applications within 5G open Radio Access Network environments. Nineteen network operators have signed on for these initial deployments, which aim to enhance 5G competitiveness.
Why it matters
By shifting from custom silicon to Nvidia's general-purpose graphics processing units, Nokia is attempting to accelerate the integration of artificial intelligence into network infrastructure. This move differentiates its product stack from competitors like Ericsson and Samsung, which do not currently support these specific user-plane applications.
The E3 interface enables dApp (decentralized application) operation with sub-millisecond latency. This setup allows Nvidia hardware to function as a direct alternative to the custom silicon traditionally used in radio access network infrastructure.
The players
Nokia
A Finnish telecommunications company that provides 5G infrastructure and develops open radio access network solutions.
Nvidia
A semiconductor company known for its high-performance graphics processing units now applied to AI-driven network software stacks.
Bell Canada
A major Canadian telecommunications provider that has conducted field trials using the Nokia hardware and software suite.
Telstra
An Australian telecommunications company that has endorsed the use of Cohere's USM technology in conjunction with these network upgrades.
Vodafone
A global telecommunications operator that has signaled its support for the integration of Cohere's USM technology within network deployments.
The details
Nokia utilizes the Nvidia Aerial framework to build software stacks that run directly on general-purpose graphics processing units. By implementing the E3 interface, the system can process dApps within the user-plane, the portion of the network responsible for carrying actual traffic. This architecture allows the platform to achieve the required sub-millisecond latency necessary for real-time radio resource management, a task previously reserved for proprietary hardware.
Timeline
September 30, 2026: Official publication of the technology integration details.
The Tech Race
Nokia's adoption of Nvidia GPUs represents a structural shift away from the custom silicon-based Radio Access Network architecture that has defined 5G hardware development. This strategy positions Nokia to compete more aggressively against incumbents like Ericsson and Samsung by utilizing a software-defined stack.
Network operators will first see these capabilities deployed in radio access networks, potentially leading to faster and more responsive data services for end users. The shift requires telecom providers to adopt the new software stack and Nvidia-based hardware, marking a significant change in core infrastructure procurement.
The takeaway
Nokia's integration of Nvidia's GPU stack suggests a broader industry trend toward commoditizing network hardware to support AI workloads. Watch for the performance results of the 19 operators currently initiating these deployments to see if general-purpose silicon can match the stability of custom-built radio hardware.
Further reading
For more on the intersection of telecom infrastructure and machine learning, see our Artificial Intelligence coverage.
Source note: This article includes information reported by ETTelecom.
Live Poll
Is now a good time for mobile providers to prioritize faster network tech over broad compatibility?
