Can Armada's Leviathan Tech Reshape AI with Clean Energy?

Armada, a hyperscale edge data centre company based in the United States, has unveiled its largest modular unit to date, Leviathan, alongside securing a US$131m strategic funding round.

Designed for megawatt-scale performance and rapid deployment, Leviathan targets sustainability by pairing high-density compute with surplus energy in remote or underutilised regions.

The funding round includes backing from new strategic partners Pinegrove, Veriten and Glade Brook, with continuing investment from Founders Fund, Lux Capital and M12, Microsoft's venture capital division.

Armada’s expansion comes amid increasing pressure to reduce the environmental footprint of artificial intelligence infrastructure, while boosting energy-efficient deployment at the network’s edge.

Targeting scale and sustainability at the edge

Leviathan extends Armada’s Galleon product line of ruggedised modular data centres (MDCs), which are pre-fabricated, transportable units engineered to deliver computing power in areas without the need for extensive traditional infrastructure.

These MDCs are optimised for edge computing, a model in which data is processed closer to the location where it is generated, reducing latency and improving energy efficiency.

Each Leviathan unit delivers megawatt-scale compute capacity and supports compute-intensive tasks such as training and running artificial intelligence models.

Compared to Armada’s Triton unit, Leviathan offers 10 times the capacity while maintaining modular flexibility.

“American energy and AI dominance hinges on one thing: moving massive compute to the edge, fast, where data and low-cost power live,” explains Dan Wright, Co-Founder and CEO of Armada.

“Leviathan, the newest member of our Galleon product line, does exactly that.

"Each unit delivers megawatt-scale performance in a fraction of the time and much more flexibly than traditional data centres, due to their ability to rapidly adapt to changes in AI chips and cooling and collocate with all available land and energy, regardless of its form or location.”

The unit’s ability to operate with stranded natural gas, solar, nuclear or other alternative energy sources allows Armada to locate infrastructure in energy-rich but digitally underserved regions.

By co-locating compute infrastructure with otherwise underutilised power, Armada aims to cut energy waste and reduce the emissions associated with long-distance data transfer.

Portable compute infrastructure designed for reuse

Leviathan is powered by the Armada Edge Platform (AEP), the company's software and control layer that manages and redeploys modular compute units.

With AEP, Leviathan can be reconfigured and moved across regions, aligning computing needs with evolving demands and available energy.

Leviathan is also designed to be "liquid-cooled", a technique that reduces the environmental impact of data centres by using fluids to remove heat more efficiently than traditional air-cooling methods.

This helps to cut both energy usage and water consumption.

“Leviathan embodies our belief that hardware should behave like software: deployable, upgradeable and governed entirely by code,” says Pradeep Nair, Founding Chief Technology Officer of Armada.

“Paired with the turnkey Armada Edge Platform and its growing Marketplace of AI models and applications, Leviathan’s liquid-cooled, energy-agnostic modules deliver megawatt-scale training and inference in weeks, bridging the digital divide by bringing AI straight to the edge, where data and power already live.”

The may reflect a growing industry preference for decentralised data centres, which are more adaptable and environmentally aligned than traditional hyperscale builds.

Modular architecture also enables companies to scale up infrastructure incrementally, helping manage both carbon output and costs.

Powered by local energy

Armada is now installing Leviathan units through partnerships with energy sector firms Fidelis New Energy and Bakken Energy.

Deployment areas include North Dakota, Texas, West Virginia and Louisiana, regions with untapped power but historically limited access to digital infrastructure.

By embedding compute infrastructure into these local energy corridors, Armada shifts AI deployment away from conventional hubs and into areas where renewable or stranded energy can be put to immediate use.

“From day one, we backed Armada because they saw what others missed: America’s AI leadership hinges on owning the entire stack, from power and silicon to software and being able to deploy it anywhere,” says Trae Stephens, Partner at Founders Fund, one of Armada's early investors.

“Leviathan drives that vision forward, expanding the Galleon line-up from suitcase-sized edge nodes for lightweight analytics and inference to megawatt-scale modules that can train and serve frontier models in the harshest environments.

"With a Galleon for every workload, Armada keeps US and allied AI efforts a step ahead.”

Armada's approach signals a change in how sustainable AI infrastructure might evolve, modular, fast to deploy and integrated with underutilised local energy sources. By decoupling data centre expansion from traditional grid and land constraints, Leviathan offers an alternative model that aligns AI deployment with energy availability rather than geography alone.