Building Sovereign AI with Azure Local: NVIDIA RTX PRO and Rubin Platform Support

Building Sovereign AI with Azure Local: NVIDIA RTX PRO and Rubin Platform Support

Governments and regulated industries face a critical challenge in AI adoption: how to harness advanced AI capabilities while maintaining strict control over data, models, and infrastructure within sovereign boundaries. Microsoft's latest announcement addresses this challenge head-on by expanding Azure Local's capabilities to support NVIDIA's cutting-edge GPU platforms.

The Sovereign AI Imperative

Across governments, regulated industries, and critical infrastructure, organizations are rapidly adopting advanced AI but with one non-negotiable requirement: AI systems must be built and operated within sovereign boundaries. The question has shifted from whether to use advanced models to where those models should run, who controls them, and how they can evolve without compromising sovereignty, resilience, or trust.

Many customers are building what Microsoft calls "Sovereign Private Clouds"—environments designed to run their most sensitive and mission-critical workloads entirely within their own operational boundaries. These environments often span on-premises datacenters, edge locations, and even fully disconnected deployments.

Azure Local: Bringing Azure Consistency to Sovereign Environments

Azure Local helps bring Azure-consistent infrastructure, lifecycle management, and governance into customer-controlled environments. As AI workloads evolve rapidly, customers need confidence that their local infrastructure can keep pace without requiring complete re-architecting.

This is particularly crucial for organizations planning for the next generation of AI acceleration while maintaining their current operations. Azure Local provides a consistent deployment, management, and governance model that spans both current NVIDIA Blackwell-based systems and future NVIDIA Rubin platforms.

NVIDIA Rubin: Powering Next-Generation Sovereign AI

The NVIDIA Rubin platform represents the next wave of accelerated computing, targeting large-scale workloads that go beyond basic inference. These include:

• Frontier model inference
• Agentic and reasoning-based systems
• High-throughput AI pipelines where performance, efficiency, and interconnect matter at scale

These are precisely the workloads Sovereign Private Cloud customers are planning for:

• National and sector-specific AI models trained and operated within sovereign boundaries
• Advanced reasoning and agentic systems for defense, intelligence, and critical decision-making
• High-performance inference pipelines operating close to sensitive data with low latency and strict compliance requirements

Foundry Local, built on Azure Local, provides a modern AI platform for deploying, operating, and scaling advanced models under full customer control while remaining aligned with Azure API, tooling, and governance.

Enterprise AI Acceleration Available Today

While Rubin represents the future of large-scale AI acceleration, Azure Local also supports enterprise-grade GPUs available today. Built on the NVIDIA Blackwell architecture, the NVIDIA RTX PRO™ 6000 Blackwell Server Edition delivers powerful AI and visual computing performance for on-premises and sovereign environments.

Azure Local harnesses the RTX PRO 6000 for scenarios such as:

• AI inferencing
• Code development
• Data-adjacent workloads

Support for the NVIDIA RTX PRO 6000 Blackwell Server Edition is generally available starting with Azure Local 2603 release. Additionally, Azure Local will soon support upcoming models in the Blackwell series, including the NVIDIA RTX PRO™ 4500 Blackwell Server Edition.

This dual approach allows customers to start building and operating sovereign AI workloads today while preparing for next-generation platforms like NVIDIA Rubin as they become available.

From Infrastructure to Models: Bringing AI Closer to Data

With Foundry Local services on Azure Local, customers can deploy and operate advanced AI models directly inside their sovereign environments. This includes access to modern models and APIs from the Foundry catalog, alongside NVIDIA-optimized models and frameworks, enabling high-performance AI execution close to sensitive data.

Using Azure Kubernetes Services (AKS) on Azure Local, customers can containerize and deploy AI workloads consistently across public and sovereign environments. This supports scenarios where models are built in the cloud and operated locally, or built and run entirely within customer-controlled boundaries, while maintaining a consistent application and operations model.

A Platform That Evolves With Customers

Azure Local is more than infrastructure. Through Azure Arc, it delivers consistent management, policy enforcement, and lifecycle operations across sovereign environments, with integrated security and monitoring capabilities.

This platform approach allows Sovereign Private Cloud customers to adopt new GPU architectures, AI frameworks, and models over time without re-architecting their environments, while maintaining strong security posture and compliance with regulatory requirements.

Microsoft is working closely with NVIDIA and OEMs including Dell, HPE, and Lenovo to enable validated hardware platforms and GPU support over time. This collaboration ensures customers can adopt new AI infrastructure with confidence as part of their Sovereign Private Cloud strategy.

The Path Forward

As AI workloads continue to advance, Azure Local remains focused on one simple goal: enabling customers to run demanding AI workloads securely, compliantly, and under their control, today and into the future.

For organizations building sovereign AI capabilities, this announcement represents a significant milestone—providing both immediate access to enterprise-grade AI acceleration through NVIDIA RTX PRO and a clear path to next-generation capabilities with NVIDIA Rubin, all within a consistent Azure-based management framework that respects sovereign boundaries.

To learn more about Azure Local and sovereign AI capabilities, visit: https://aka.ms/azurelocal