Super Micro Computer Partners With Nokia To Deploy AI-RAN From Distributed Cell Sites To Centralized Network Hubs

The solution combines Nokia's RAN technology and network expertise with the NVIDIA Aerial RAN Computer platform, supported by Supermicro's AI-optimized infrastructure. By bringing accelerated compute closer to the access network, it enables radio and AI workloads to run on shared infrastructure, supporting applications which require proximity and predictable service levels-including voice AI, vision AI, physical AI, and internal uses such as autonomous network operations.

AI RAN Solutions for the Network Edge

Embedding accelerated compute into the RAN enables operators to run network and AI workloads on the same infrastructure, improving spectral efficiency, automating network operations, and optimizing energy use while creating new opportunities for edge AI services. Combining Supermicro systems with Nokia's RAN technology and software, the NVIDIA Aerial RAN Computer platform provides modular, right-sized CPU and GPU configurations for deployment from cell sites to MSO and aggregation locations, with NVIDIA ConnectX technology providing high-speed, software-defined networking. Together, these capabilities give operators a flexible foundation for low-latency AI applications today and a software-driven evolution path towards 6G networks.

Supermicro's modular, right-sized systems for NVIDIA ARC span distributed cell sites, aggregation locations, and core hubs, enabling operators to precisely deploy AI capacity where it is needed.

Distributed cell sites: The ARC MGX 2U Short-depth platform brings inference closer to radios, users, and applications. NVIDIA MGX is a modular reference architecture for accelerated computing, helping system builders design future-compatible systems faster, reducing costs, accelerating time to market, and improving ROI. Supermicro's dual-node ARC-Pro combines an NVIDIA Grace C1 CPU and NVIDIA RTX PRO 4500 Blackwell Server Edition GPU per node in a 429 mm chassis, delivering up to 1.6 PFLOPS of FP4 AI performance. For edge networking functions such as Service Management and Orchestration, the ARC-Compact delivers high efficiency compute performance in a 1U, short-depth form factor.
Centralized MSO and aggregation sites: The ARC MGX 2U supports higher-throughput, coordinated inference. This system class comes with a choice between dual NVIDIA Vera CPUs or dual x86 processors and two NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. These systems support multi-cell workloads, including coordinated scheduling and beamforming, as well as voice, vision, and physical AI.
Core network hubs: The ARC MGX 4U provides high-density, air-cooled inference with up to eight NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. Supermicro's 4U AI servers deliver dual x86 CPUs and up to 6 TB of DDR5 memory to support large-scale inference and model update workloads. This configuration comes with the option for an NVIDIA ConnectX-8 switchboard, providing PCIe Gen 6 for even faster GPU-GPU interconnectivity.
Supermicro's servers for NVIDIA ARC are generally available now. The ARC MGX 2U with dual Vera processor is on display on Supermicro's booth at NVIDIA GTC Berlin, October 20–22, 2026, and will be available later this year. The 2U ARC-Pro is planned for customer testing later this year, with general availability in the first half of 2027.