Armada and Nscale have signed a letter of intent to scale sovereign AI infrastructure across global markets, targeting both centralized and edge deployments for public sector and enterprise customers. Through the agreement, the two companies plan to combine large-scale AI capacity with rapidly deployable modular infrastructure.

The collaboration brings together complementary capabilities. Nscale, headquartered in Europe, develops high-performance AI infrastructure spanning power, data centers, compute, and software. Meanwhile, San Francisco-based Armada focuses on distributed intelligence through modular data centers, known as Galleons, and its proprietary Armada Edge Platform.

Full-Stack Infrastructure From Core to Edge

Under the proposed framework, the companies intend to deliver full-stack AI solutions at multiple locations worldwide. These offerings are expected to include modular data centers, GPU compute, application software, and operational support.

In addition, access to land and power at selected sites has already been identified. As a result, deployments can be accelerated for customers in both government and private sectors.

By combining these assets, sovereign compute environments can be established in regions where traditional infrastructure is limited or absent. In many cases, delivery timelines are expected to be significantly shorter than those of conventional data center builds.

Hub-and-Spoke Model Supports Global Reach

The partnership is structured around a hub-and-spoke architecture. Large-scale data centers developed by Nscale will provide foundational capacity and cost efficiency. At the same time, Armada’s modular systems will extend sovereign AI capabilities closer to end users.

For example, Armada’s Leviathan Galleon, a megawatt-scale modular deployment, is designed to bring high-performance compute to edge locations. These systems allow AI workloads to be operated securely while maintaining local control.

Consequently, customers can balance centralized capacity with distributed execution across geographies.

Demand for Sovereign AI Accelerates

Executives from both companies pointed to rising demand for operational AI that meets sovereignty and compliance requirements. According to Nscale founder and CEO Josh Payne, enterprises and governments increasingly require infrastructure that can scale while remaining secure and locally governed.

Similarly, Armada co-founder and CEO Dan Wright said organizations now expect AI infrastructure to extend beyond centralized clusters. He noted that modular deployments allow sovereign AI to be delivered with greater speed and flexibility.

Toward a Repeatable Global Model

Looking ahead, Armada and Nscale plan to develop a repeatable model for deploying sovereign AI infrastructure worldwide. By aligning large-scale cloud platforms with modular compute and distributed operations, AI adoption can be accelerated without compromising security or performance.

If finalized, the partnership would position both companies to support sovereign AI deployments across a wide range of environments, from national data centers to edge locations serving emerging markets.