4 Power Plays as AI Data Centers Demand $7 Trillion in Cooling

The explosive growth of artificial intelligence is creating a secondary boom for a less glamorous but equally critical sector: the industrial companies that keep data centers from melting. As AI workloads demand unprecedented levels of electricity, the market for power management and advanced cooling solutions is surging, turning companies like Eaton and Vertiv into essential suppliers for the AI revolution. Analysts project that global spending on data center buildouts could exceed $7 trillion by 2030, with a substantial portion dedicated to upgrading the power infrastructure that underpins them.

"This transaction commercializes the first building of our newest gigawatt-scale campus and marks our second AI data center lease," Asher Genoot, CEO of Hut 8, said in a statement regarding a new Texas facility. "More importantly, it demonstrates that our development model, which pairs power-first underwriting with disciplined commercialization and institutional execution, is repeatable and extendable across our broader pipeline."

The scale of these new projects is immense. Hut 8's Beacon Point campus in Texas, for example, secured a 15-year, $9.8 billion lease to provide 352 megawatts (MW) of IT capacity for a single high-investment-grade tenant. This one deal brings Hut 8's total contracted AI data center capacity to 597 MW. Power management giant Eaton saw orders in its Electrical America segment jump 200 percent last year, contributing to a backlog of nearly $20 billion.

This spending spree is not just about adding more servers; it's a fundamental overhaul of the electrical grid and data center architecture. The sheer power density of modern AI accelerators, like Nvidia's latest GPUs, puts immense strain on an aging U.S. electrical grid. This has created a "grid-to-chip" opportunity for companies that can manage every link in the power supply chain, from utility-scale transformers down to the server rack cooling systems.

The Grid-to-Chip Overhaul

Eaton (ETN) is a primary beneficiary of this trend, with its stock rising 30 percent in 2026. The company manufactures a vast portfolio of electrical components, including switchgears, transformers, and circuit breakers, that are essential for modernizing the grid. U.S. utility companies are expected to invest $1.4 trillion in grid infrastructure by 2030 to handle the increased load from AI and broader electrification.

To capture more of the value chain, Eaton acquired Boyd Thermal for $9.5 billion last year. The deal brings specialized liquid cooling and thermal management technologies in-house, a segment Eaton expects to grow by 35 percent annually over the next three years. This positions the company as a critical supplier for both the power delivery and heat removal aspects of AI data centers.

A New Breed of Data Center

The Hut 8 (HUT) deal in Texas exemplifies the new, partnership-driven model for building AI infrastructure at speed and scale. The project brings together a roster of industry leaders: Nvidia is the technology partner, with the 352 MW facility engineered to its DSX reference architecture for AI factories. Vertiv (VRT) is supplying the critical digital infrastructure systems, including power and cooling, while Jacobs is leading the engineering and construction.

This repeatable delivery model is crucial for hyperscalers who need to deploy massive amounts of compute capacity quickly. For its part, Vertiv is working alongside Hut 8 to converge power, cooling, and deployment execution. "Next generation AI infrastructure will be defined by how quickly power can be converted into AI capacity," said Giordano Albertazzi, CEO of Vertiv.

The investment thesis for these industrial suppliers is straightforward: while competition among AI model providers and chip designers is fierce, the underlying need for power and cooling is universal. Eaton, which has paid a dividend every year since 1923, offers a "picks-and-shovels" play on the AI boom, benefiting from the buildout regardless of which AI company ultimately wins. While risks remain, including potential public pushback on the environmental impact of data centers, the demand for AI compute shows no signs of slowing. For investors, the companies powering the revolution may offer a more durable path to growth than the high-flying tech stocks they support.

This article is for informational purposes only and does not constitute investment advice.