Energy Vault and Crusoe Forge Pact to Power the AI Revolution

SNYDER, TX – February 11, 2026 – In a strategic move to address the voracious energy and infrastructure demands of the artificial intelligence boom, energy storage leader Energy Vault and AI infrastructure provider Crusoe have announced a landmark agreement to build the next generation of data centers. The partnership will see the deployment of Crusoe’s modular “Spark” AI factories at Energy Vault’s technology site in Snyder, Texas, beginning in 2026.

This collaboration aims to slash the time it takes to bring powerful AI computing capacity online, directly tackling the critical bottlenecks that currently threaten to stall the pace of AI development. The initial phase will support up to 25 megawatts (MW) of power load, a significant footprint designed to expand Crusoe’s AI cloud services and meet escalating customer demand. The initiative represents a new blueprint for how the energy and tech sectors can co-evolve, pairing rapid-deployment energy systems with modular, high-density compute.

Addressing the AI Infrastructure Bottleneck

The artificial intelligence industry is expanding at a breakneck pace, but its growth is tethered to a stark physical reality: an insatiable need for electricity and the data centers to house the powerful hardware. The global AI infrastructure market, which stood at over $80 billion in mid-2025, is projected by analysts to skyrocket past $750 billion by 2029. This exponential demand is placing unprecedented strain on power grids and traditional data center construction models.

Global data center electricity consumption is on track to double by 2030, with AI workloads being the primary driver. An AI-focused data center can consume as much electricity as tens of thousands of homes, and the lead time for securing power and building new facilities can stretch for years. This has created a critical chokepoint where the ambition of AI innovation is colliding with the limitations of physical infrastructure. Power availability, grid interconnection queues, and the sheer complexity of building facilities for GPU racks that can draw over 100 kW each have become the biggest hurdles to growth.

“Crusoe is executing against a clear market imperative: customers need scalable compute delivered quickly and reliably,” said Robert Piconi, CEO and Chairman of Energy Vault, in a statement announcing the deal. The partnership is designed to circumvent these long timelines by fundamentally rethinking how data centers are powered and built.

A Modular Solution for an Exponential Market

The core of the Energy Vault-Crusoe strategy lies in modularity and vertical integration. Instead of a traditional, multi-year construction project, the companies will deploy a system of interlocking, prefabricated components. Energy Vault will provide the foundational “powered shell” infrastructure, a standardized and repeatable system that includes all the necessary electrical and mechanical hardware to support high-density computing.

This powered shell is designed to be rapidly deployed on-site, creating a ready-made landing zone for Crusoe’s proprietary Spark units. These units are, in effect, compact, factory-built AI data centers, fully equipped with high-performance computing and cooling systems. By manufacturing the Spark units off-site while Energy Vault prepares the powered site, the two companies aim to materially reduce delivery timelines and enable a 'plug-and-play' model for scaling capacity as customer demand grows.

This integrated approach leverages the strengths of both companies. Energy Vault brings its expertise in engineering and operating mission-critical energy systems, supported by its proprietary energy management software. Crusoe, which positions itself as an “AI factory company,” contributes its vertically-integrated model, designing and manufacturing its own hardware to ensure quality and consistency for its Crusoe Cloud platform.

“By manufacturing Spark units and partnering with the talented team at Energy Vault, we are furthering Crusoe’s vision of vertically-integrated, energy-first AI infrastructure,” stated Cully Cavness, Co-Founder and President of Crusoe. He emphasized the model’s flexibility for various applications, from large training clusters to distributed, low-latency inference deployments.

Energy Vault’s Strategic Pivot to High-Margin AI

For Energy Vault, this agreement marks a significant strategic expansion beyond its core business of grid-scale energy storage. The move catapults the company into the heart of the high-growth AI infrastructure market and enhances its “Own & Operate” business model, where it builds and manages assets to generate long-term, recurring revenue.

Most notably, the financial implications are transformative. Energy Vault projects that its AI infrastructure deployments will deliver a step-change in profitability, yielding an estimated 10 to 20 times higher EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) per megawatt compared to its traditional energy storage projects. This dramatic increase is driven by the combination of long-term, contracted cash flows from AI clients, the cost efficiencies of a standardized modular design, and the premium market for rapidly deployable compute capacity.

This venture allows Energy Vault to capitalize on its core competencies in energy management and infrastructure deployment while capturing the significantly higher margins available in the AI sector. The partnership with Crusoe serves as a commercial platform that the company plans to replicate, accelerating its overall growth and establishing a powerful new revenue stream tied directly to the AI revolution.

Texas Emerges as a Next-Generation Compute Hub

The choice of Snyder, Texas, for the initial deployment is no coincidence. The state’s energy-rich landscape, including abundant renewable resources and a robust power grid, is making it a magnet for energy-intensive industries, chief among them data centers. This project highlights a growing trend where AI infrastructure is being co-located with power generation, creating synergistic ecosystems that can operate more efficiently.

Crusoe’s “energy-first” philosophy, which includes innovative methods like harnessing otherwise stranded or flared natural gas to power its operations and reduce methane emissions, aligns well with the complex energy environment of a state like Texas. As more AI data centers come online, their immense and volatile power demands will require sophisticated energy management and storage solutions to ensure grid stability. The Energy Vault-Crusoe model, which integrates energy infrastructure directly with compute, provides a potential blueprint for how regions can attract high-tech investment while managing the impact on their power resources.

This deployment not only promises local economic benefits but also positions the region at the forefront of next-generation infrastructure development. As the AI industry continues its relentless expansion, the ability to rapidly deploy reliable, power-aware compute capacity will be a key competitive advantage, and partnerships like this one are paving the way for that future.