Armada, a startup focused on edge AI infrastructure, just raised $230 million in Series B funding at a $2 billion valuation to build the first dedicated factory for modular AI data centers in the United States. The company's partnership with Johnson Controls represents a fundamental shift in how America approaches AI infrastructure: instead of relying on one-off deployments or imported solutions, the country is now investing in domestic manufacturing capacity to produce AI computing systems at scale.

Why Is a Data Center Factory Such a Big Deal?

The Galleon Forge One facility, a 400,000-square-foot factory in Arizona, will begin continuous production in summer 2026 and is expected to create over 500 jobs in manufacturing and supply chain roles. This is not just another data center announcement. It represents the first major effort to industrialize modular AI infrastructure production in the United States, moving beyond prototype deployments to assembly-line manufacturing.

Dan Wright, co-founder and CEO of Armada, framed the stakes clearly: "The AI race will not be won by one-off projects. It will be won by the companies and countries that can manufacture, deploy, and continuously improve AI infrastructure, with speed, scale and sovereignty." This reflects a growing recognition among U.S. policymakers and investors that AI dominance depends not just on chip design or software, but on the ability to physically build and deploy computing infrastructure faster than competitors.

What Makes Armada's Infrastructure Different?

Armada's modular data centers, called Leviathan, are designed to be deployed in environments where traditional data centers cannot operate. The company has already demonstrated real-world deployments across multiple demanding use cases:

• Defense Applications: Armada supported an allied defense organization in deploying a Triton unit into the field in just six days, bypassing traditional data center construction timelines.
• Renewable Energy Integration: WinDC selected Armada to build portable AI factories designed to run on renewable energy that Australia's national grid cannot absorb, starting with a Triton and scaling to a 10-megawatt-plus Leviathan deployment.
• Maritime Operations: The U.S. Navy deployed Armada's edge computing during a multinational maritime exercise, running mission applications on a Galleon aboard a communications-constrained vessel to enable real-time decision-making.
• Offshore Energy: Aker BP is testing Armada's infrastructure on the Norwegian Continental Shelf, with plans to deploy a Galleon on the Deepsea Nordkappa rig to enable autonomous operations across its fleet.

These deployments reveal why modular infrastructure matters: traditional data centers take months or years to build and require stable power grids, fiber optic connections, and cooling infrastructure. Armada's systems can be deployed in days or weeks, operate in remote locations, and scale incrementally as demand grows.

How Does Johnson Controls Strengthen Armada's Position?

Johnson Controls, a 140-year-old building systems company, brings critical expertise in thermal management and mission-critical infrastructure. The company operates more than 40,000 field personnel across key regions globally and has deep manufacturing capabilities in the United States. This partnership is strategic: cooling is one of the most expensive and complex aspects of data center operations, and Johnson Controls' expertise directly addresses this bottleneck.

"Johnson Controls is working with Armada to rapidly deliver secure modular data centers at scale. Together, we have already deployed units across the United States and around the world, demonstrating the expertise and global reach required to support mission-critical environments," said Joakim Weidemanis, Chief Executive Officer of Johnson Controls.

Joakim Weidemanis, Chief Executive Officer at Johnson Controls

The partnership also signals confidence from a major industrial player that modular AI infrastructure is not a niche market but a core business opportunity. Johnson Controls' investment in Armada, combined with its manufacturing and service capabilities, suggests the company sees long-term demand for this infrastructure type.

What Do the Numbers Tell Us About Market Demand?

Armada's growth metrics are striking. From fiscal year 2025 to 2026, the company recorded 540 percent customer bookings growth, with significant interest in its Leviathan systems. In the first quarter of fiscal year 2027 alone, bookings grew 2,000 percent compared to the same quarter the previous year. These numbers suggest that demand for modular, deployable AI infrastructure is accelerating rapidly across industries and geographies.

The Series B funding round was heavily oversubscribed, bringing total funding to nearly half a billion dollars. The investor syndicate includes Overmatch, BlackRock, 8090 Industries, NightDragon, Mitsui, and Singtel Innov8, alongside existing backers like Felicis, Marlinspike, and Founders Fund. The participation of BlackRock, one of the world's largest asset managers, signals institutional confidence that AI infrastructure manufacturing is a durable investment thesis, not a speculative bubble.

What Does This Mean for the Broader AI Infrastructure Market?

Armada's factory announcement arrives at a critical moment. The AI industry has been constrained by two bottlenecks: semiconductor availability and power grid capacity. While chip shortages have eased somewhat, power infrastructure remains the binding constraint on AI deployment. Modular data centers that can operate independently of centralized grids address this constraint directly.

The emphasis on "sovereignty" in Armada's messaging also reflects geopolitical concerns. By manufacturing AI infrastructure domestically and enabling customers to deploy it wherever they need it, the company positions itself as a solution to concerns about data residency, supply chain vulnerability, and technological independence. This resonates with defense organizations, allied governments, and enterprises handling sensitive data.

The Galleon Forge One factory represents a bet that the future of AI infrastructure is not centralized hyperscale data centers owned by a handful of cloud providers, but distributed, modular systems deployed closer to where data is generated and where customers need real-time processing. If this thesis proves correct, Armada's manufacturing capacity will become increasingly valuable as demand for edge AI computing grows.