Meta, SpaceX fuel industrial boom: Caterpillar backlog hits record $63B

In October 2025, Meta announced plans to construct a 2-gigawatt data center. By July 2026, the company had upgraded that facility's planned capacity to 5 gigawatts, signaling the intensity of demand for AI computing infrastructure. Meta is not alone: Space Exploration Corporation is leasing out AI computing power through facilities called Colossus I and Colossus II. The rush to build these massive data centers is creating a substantial business opportunity for industrial equipment manufacturers previously unconnected to the AI sector.

Caterpillar, a maker of earth-moving equipment and on-site power generators, is among the primary beneficiaries. The company's products are essential for constructing data centers and addressing one of the sector's key pain points: the enormous electricity demands of these facilities strain the electrical grid. On-site power generation that does not rely on grid capacity has become a major selling point. In the first quarter of 2026, Caterpillar reported revenues up 22% and adjusted earnings higher by 30%, direct evidence of the AI infrastructure tailwind. The most striking figure is the company's backlog, which reached a record $63 billion（約10兆円）—a 79% increase compared to Q1 2025. This backlog essentially represents future revenue already committed, providing clear visibility into sustained demand over the coming quarters.

Eaton, which manufactures electrical products for power management that form the foundation of AI data centers, is also capturing significant benefits. The company's Americas business division has seen its backlog grow by over 40% year-over-year, with order growth expanding at a 60% rate, both driven by AI demand. The article notes that Eaton has been divesting older business units to streamline operations, which complicates year-over-year comparisons, but the strong backlog provides what the article calls "a fairly clear line of sight for future revenues and earnings."

The investment case for both stocks carries a key caveat: Wall Street has already bid up valuations substantially in anticipation of the AI build-out. Caterpillar trades at a price-to-earnings ratio of 45×, while Eaton's stands at 38×—both figures well above historical norms and unappealing to value-conscious investors. However, the article suggests the data center construction race "doesn't look anywhere near over," implying the underlying demand dynamics remain strong enough to potentially justify continued investment, provided one believes AI infrastructure spending will persist.

The race to build massive AI data centers—exemplified by Meta's 5-gigawatt facility and SpaceX's Colossus leasing operations—is creating an unexpected tailwind for traditional industrial manufacturers. Caterpillar and Eaton, companies that operate far outside the consumer tech spotlight, are positioned directly in the supply chain for this infrastructure boom. Caterpillar supplies the heavy construction equipment required to build these "massive structures," as well as on-site power generators that address the grid-strain issue created by data center electricity demands. Eaton's electrical power management products form the backbone of data center infrastructure. The body indicates that both companies face a similar dynamic: strong backlogs (Caterpillar at record levels, Eaton up over 40%) signal that orders are already flowing in, making future revenues predictable. However, the article notes a valuation concern: Wall Street's intense focus on AI has already driven up both stocks' multiples—Caterpillar to 45× earnings and Eaton to 38×—making them unattractive to value-focused investors despite the underlying strength of their order books. The article suggests the data center build-out is still in early innings, implying the demand tailwind may persist, but the elevated entry valuations complicate the investment case.