1Finity puts mobile networks and AI on shared computing infrastructure
1Finity has developed software that places mobile radio-network functions and AI workloads on shared NVIDIA computing infrastructure. The Fujitsu group company plans a December launch with compatible radio equipment and field validation with operators in fiscal 2027. Its proposed enterprise services combine private wireless connections, computing and AI applications, including uses that need stable communications and real-time control.
Artificial Intelligence··Morning
Radio functions share computing with AI
1Finity, the Fujitsu group company that develops communications networks, announced Open vRAN Solutions VB5000 on November 4. The software places radio access network functions and AI workloads on the same NVIDIA graphics-processor infrastructure. A radio access network connects wireless devices with an operator’s core network. The company plans to offer the software with compatible radio equipment globally from December.[1], [2]
The product is a core part of 1Finity’s AI-Native vRAN approach, combining virtualised radio access with AI services. Operators would be able to offer enterprises private wireless networks, computing resources and AI applications together. The company describes shared infrastructure as a way to avoid having to build separate computing environments for the communications functions and the AI work.[1]
Operators choose infrastructure and application partners
The proposed infrastructure combines NVIDIA AI Aerial, the chip company’s platform for radio and AI computing, with servers from Quanta Cloud Technology and Supermicro. 1Finity plans to assemble the components around individual operators’ requirements. Its radio equipment follows Open Radio Access Network specifications, known as O-RAN, which address connections between radio and base-station components supplied by different vendors.[1]
Armada’s Bridge software can be added for workload coordination, billing and access security where several organisations share an environment. Aible’s no-code application development platform is another supported component. Its templates cover the development, deployment and operation of AI agents. These software partners provide functions above the shared infrastructure that runs both the communications and AI workloads.[1]
Operator field validation is planned for fiscal 2027
1Finity plans field validation of the AI-Native vRAN approach with mobile operators during fiscal 2027. It intends to use the findings in later network development. Enterprise applications include robots and other systems that require real-time control and stable communications. The announced path combines the December product introduction with subsequent operator testing, as the company works on offering mobile connectivity and AI services through a common platform.[1]
