AI Data Center Boom Creates Windfall for Industrial Suppliers Cat, Eaton

The artificial intelligence boom is reshaping the industrial economy in ways that have little to do with algorithms or neural networks. While tech giants race to build the massive computing facilities that will power the next generation of AI systems, a quieter but equally significant story is unfolding in the supply chains that make those buildings possible. Two unglamorous industrial companies—Caterpillar and Eaton—are experiencing a windfall that reveals how the digital revolution depends entirely on physical infrastructure.

Meta's ambitions illustrate the scale of what's happening. In October 2025, Mark Zuckerberg announced plans for a data center capable of handling 2 gigawatts of power. Nine months later, by July 2026, that same facility had been redesigned to handle more than double that capacity—5 gigawatts. Meta is not alone in this expansion. Space Exploration Corporation is building what it calls Colossus I and Colossus II, massive computing complexes that it leases out to companies hungry for AI processing power. These are not modest undertakings. They are among the largest industrial construction projects in the world, and they require equipment, materials, and expertise that only a handful of companies can provide.

Caterpillar, the heavy equipment manufacturer, sits at the center of this demand. The company makes the bulldozers, excavators, and graders needed to prepare sites and move earth on a scale that matches the size of these data centers. But Caterpillar also manufactures something equally critical: on-site power generators. As data centers consume electricity at rates that strain regional power grids, companies are increasingly turning to self-contained power solutions that don't depend on the public utility system. For Caterpillar, this has meant a dramatic surge in orders. In the first quarter of 2026 alone, the company's revenues climbed 22 percent compared to the prior year, while adjusted earnings jumped 30 percent. But the most telling number is the company's backlog—the orders it has already received but not yet fulfilled. That backlog now stands at $63 billion, up 79 percent from the first quarter of 2025. In practical terms, this represents future revenue that is already locked in, a cushion of certainty that most companies would envy.

Eaton, which manufactures the electrical infrastructure that forms the backbone of data center operations, is experiencing a similar surge. The company's Americas division has seen its backlog grow by more than 40 percent year over year, driven almost entirely by demand from AI data center projects. More striking still is the rate at which new orders are coming in—order growth has accelerated at a 60 percent pace. Eaton has been consolidating its business, divesting older divisions to focus on higher-growth areas, which complicates the financial picture somewhat. But a swelling backlog provides something that investors and analysts value highly: visibility into future revenue. When a company has orders on the books, it knows roughly how much money it will make in the coming quarters.

What makes this story significant is not just the size of the numbers, but what they reveal about the relationship between the digital and physical worlds. The companies building artificial intelligence systems need to construct actual buildings, pour concrete, run electrical cables, and install generators. They cannot do this work themselves. They depend on industrial suppliers whose business models were established decades before anyone imagined AI data centers. Caterpillar and Eaton are not technology companies, and they do not need to be. They are benefiting from the AI boom not because they understand machine learning, but because they understand how to move earth and manage electrical power at scale. As long as hyperscalers continue to build larger and more numerous data centers, these industrial suppliers will have work.