Supermicro Expands Support for AI-RAN and Sovereign AI Solutions to Deliver High-Performance and Scalable AI Data Center Infrastructure 

Cutting-edge infrastructure solutions from Supermicro feature the latest CPUs and GPUs, power AI-native telecom networks, and secure, scalable AI factories

Supermicro, Inc. (NASDAQ: SMCI), a Total IT Solution Provider for AI/ML, HPC, Cloud, Storage, and 5G/Edge, has announced expanded support for infrastructure solutions powering sovereign AI platforms and Artificial Intelligence – Radio Access Networks (AI-RAN). At MWC Barcelona, Supermicro and its ecosystem partners featured the latest real-world use cases that bring together performance, efficiency, and scalability.

“Delivering AI to the RAN at scale requires infrastructure optimized for telecom networks,” said Charles Liang, president and CEO of Supermicro. “As operators embed intelligence across their networks and advance sovereign AI strategies, Supermicro’s flexible Data Center Building Block Solutions® (DCBBS) and deep ecosystem collaborations enable rapid deployment of high-performance, energy-efficient solutions that help ensure data sovereignty and long-term scalability.”

The adoption of AI in telecom networks is growing rapidly as operators seek greater efficiency and automation. AI-RAN embeds intelligence into the Radio Access Network to optimize spectrum, energy, and performance, while simultaneously creating a distributed network that powers AI solutions for operators and end users.

AI-RAN requires infrastructure solutions that are designed to meet the unique requirements of telecom networks in a scalable, cost-efficient way. Supermicro is expanding its broad range of AI-RAN systems, in line with NVIDIA Aerial RAN Computer (ARC) design, to support the latest technologies including NVIDIA Blackwell architecture, with upcoming products using the latest NVIDIA ARC-Pro.

• ARS-111L-FR is a short-depth, 1U system designed for distributed RAN workloads. The system features NVIDIA Grace™ CPU C1 and is compatible with NVIDIA ConnectX Ethernet® adapters for high-speed networking. Additionally, the server has space for up to 2 low-profile GPU cards, such as the NVIDIA L4 for AI workloads.
• ARS-221GL-NR is a high-performance 2U Grace Superchip system with up to 2 double-width NVIDIA GPUs. Supermicro has improved the system to support NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, enabling it to run demanding AI workloads throughout the telecom network.
• ARS-111GL-NHR is Supermicro’s 1U GH200 Grace Hopper™ Superchip system with integrated GPU and 900GB/s NVLink® interconnect between CPU and GPU. The high-speed chip-to-chip bandwidth together with support for NVIDIA BlueField® data processing units (DPUs) and ConnectX Ethernet adapters makes NVIDIA Grace Hopper a powerful and versatile system for research as well as AI inferencing at the network edge.

Sovereign AI refers to an enterprise’s ability to develop, deploy, and govern artificial intelligence within its own controlled infrastructure, data environments, and compliance frameworks. Increased demand for sovereign AI platforms represents a strategic opportunity for telecom operators, by offering secure, in-country AI infrastructure as a service, and unlocking new revenue streams in the digital economy.

Delivering sovereign AI at scale requires infrastructure that is powerful, scalable, and energy efficient. Supermicro’s innovative DCBBS are designed to enable organizations to rapidly deploy and expand AI data centers. Modular architectures simplify scaling, while advanced thermal and power designs improve operational efficiency.

At MWC, Supermicro collaborated with leading industry players to demonstrate real-world deployments and practical use cases that advance AI acceleration.

• Telenor, a leading Nordic telecom provider, has launched Telenor AI Factory, Norway’s first sovereign AI cloud platform. Designed for enterprises, startups, and public sector organizations, it provides NVIDIA GPU-accelerated infrastructure entirely within Norway, ensuring data never leaves national borders. By combining security, scalability, and sustainability, Telenor AI Factory lowers barriers to AI adoption, shortens time-to-market, and supports sensitive data compliance. This positions Telenor not just as a telecom operator, but as a strategic provider of sovereign AI services.
• SK Telecom, South Korea’s leading telecom provider, has built the Haein Cluster, a sovereign AI infrastructure platform designed for large-scale AI workloads and national competitiveness. The cluster features over 1,000 Supermicro AI servers equipped with NVIDIA Blackwell GPUs, forming one of the country’s highest-performance AI GPU clusters. Hosted at SK Telecom’s Gasan AI Data Center, it delivers GPU-as-a-Service for training, inference, and model development. Haein supports enterprises and national AI initiatives securely within South Korea, expanding service opportunities.
• Nokia, global leader in telecommunications, network infrastructure, and technology innovation, is showing several solutions based on Supermicro hardware. AnyRAN is Nokia’s software suite designed to give mobile operators and enterprises flexibility in building radio access networks (RAN) by allowing them to deploy cloud-native software on any infrastructure. Supermicro’s 1U Grace Hopper system ARS-111GL-NHR has been validated to run AnyRAN. Additionally, Nokia is introducing a validated infrastructure design for hybrid AI/ML training and inference based on Supermicro AS -8125GS-TNMR2 servers.
• Viettel High Tech, Vietnam’s top 5G network infrastructure solution provider, has built a large 5G network based on Open RAN standards. The adoption of open, disaggregated infrastructure has enabled Viettel High Tech to speed up the implementation of new technologies and applications, in addition to delivering substantial cost savings. One of the systems used in Viettel High Tech ‘s infrastructure is the Supermicro short depth 1U edge server. These systems provide flexibility, performance, and innovation, ensuring optimal support for 5G and advanced wireless networks.

More information about Supermicro Systems Optimized for AI Applications