Some of the biggest financial winners from the AI boom aren’t making chips at all. Generators, circuit breakers, and industrial turbines are becoming some of America’s most sought-after equipment, and the century-old manufacturers behind them are cashing in.
Power demand rewrites manufacturing's growth story
US manufacturing hit its highest level since 2022 last month, driven almost entirely by the AI data center buildout. Caterpillar, Cummins, and Eaton are among the industrial giants pivoting hard to capture that demand.
Caterpillar's electricity generator segment has become the company's leading source of profit, surpassing its legacy construction equipment business. The company posted a record $20.5B in quarterly revenue and saw operating profit rise 50% year over year.
Price increases alone contributed nearly $600M to that operating profit figure. Caterpillar shares are up roughly 50% since the start of 2026, and its backlog of generator orders from data centers now stretches to 2030.
Cummins is investing $450M to expand generator production, following a $200M expansion completed last year. The company expects data-center-related sales to climb 80% to $9B by 2030, up from 2026 levels.
Starting in 2028, Cummins plans to offer generators large enough to serve as a primary power source rather than just a backup. Each unit will run on natural gas and produce four megawatts of electricity from a 130-liter engine.
Eaton has spent $13B on acquisitions since 2025 to bulk up its data center business. Data centers and distributed IT now account for 21% of Eaton's sales, up from 14% at the end of 2023. The company's overall second-quarter sales rose 21% year over year.
Ford uses EV battery excess to chase data center storage
Not every company pivoting toward AI infrastructure is a traditional industrial name. Ford Motor is redirecting its surplus EV battery capacity toward energy storage for data centers through a new subsidiary called Ford Energy.
The automaker plans to spend $2B on the effort. EV demand has slumped, leaving Ford with more battery production capacity than it needs for vehicles. The data center pivot lets Ford monetize that excess rather than idle it.
The scale of investment is raising alarms
The AI compute market was valued at roughly $215B in 2025 and is projected to grow dramatically through 2040.
Hyperscalers including Google, Meta, Microsoft, and Amazon are spending between $750B and $800B annually on AI infrastructure, with some forecasts pushing that figure toward $1T, according to NFJ Investment Group's chief investment officer.
Nvidia recently backed $105B in financing for an OpenAI data center in Ohio, supporting an initial 4.25 gigawatts of computing capacity.
SB Energy will build and manage the facility on a 20-year lease, with power infrastructure investment of at least $4.2B committed to regional grid upgrades.
"Compute is really becoming the new oil, the new limited resource of the AI age."
Greg Brockman, OpenAI
That scale of spending is drawing scrutiny. Virginia recently became the first state to tax data centers based on electricity consumption, generating up to $600M in projected annual revenue.
New York imposed a one-year moratorium on data center construction. Ohio, Illinois, Arizona, and Nebraska have all paused or scaled back tax incentives for the industry.
The boom-and-bust risk
Analysts and executives are watching for signs of oversupply. Harvard professor Willy Shih has compared today’s frenzy to the late-1990s fiber boom, when massive capacity sat unused for decades.
Morningstar's industrials analyst notes that cooling and power equipment companies like Vertiv generate over 80% of revenue directly from data centers, making them highly exposed to any pullback.
Eaton, by contrast, derives only about a quarter of its sales from data centers, with the rest tied to commercial buildings and broader grid infrastructure. Cummins is expanding generator capacity in smaller increments and limiting new plant construction to avoid overbuilding.
The ISM reported recently that 20% of manufacturing GDP showed contraction last month, up from 5% in June, a signal that cost pressures are beginning to bite in parts of the industrial economy outside of AI-linked sectors.