The U.S. artificial intelligence boom is no longer confined to technology companies and massive data centers. Its impact is spreading deep into America’s manufacturing and industrial supply chains, creating a surge in demand for generators, transformers, cooling equipment, steel components, cables, pipes, construction machinery and other infrastructure needed to power the AI economy.
At the center of the trend is the rapid expansion of U.S. data centers, which require enormous amounts of electricity and specialized infrastructure. The resulting demand is giving manufacturers across multiple industries a new source of growth — even as other parts of the U.S. economy remain under pressure.
One of the clearest examples is Generac, the Wisconsin-based manufacturer best known for residential backup generators. The company is investing $250 million by the end of 2027 to upgrade multiple factories so they can produce larger generators designed for data centers.
Generac’s data-center generator backlog has already reached approximately $1.6 billion, and the company expects to add about 1,000 employees, equivalent to roughly 10% of its workforce.
But Generac is only one piece of a much larger industrial chain.
AI Demand Is Reaching Far Beyond Generators
Data centers need much more than servers and advanced chips. They require electricity generation, transformers, switchgear, cooling systems, construction materials, backup power, buildings and extensive electrical infrastructure.
That means manufacturers supplying those industries are also benefiting.
Timken, an Ohio-based producer of engineered steel bearings, said data-center projects are adding another source of demand alongside its traditional aerospace and defense customers. The construction of massive facilities, roads and power infrastructure requires a wide range of industrial products.
Smaller manufacturers are seeing the impact as well. Southeastern Hose, a Georgia-based company that traditionally supplied steel and petrochemical industries, has seen data-center demand surge. The company said its revenue has tripled over the past five years and that it has added 60 workers.
The effect is essentially a ripple: a new data center creates demand for electrical equipment; those manufacturers need components and raw materials; their suppliers then need more production capacity of their own.
U.S. Manufacturing Is Showing Signs of an AI-Driven Lift
The data suggests that AI infrastructure is contributing to a broader improvement in selected areas of U.S. manufacturing.
U.S. factories added 5,000 jobs in July, bringing manufacturing employment gains for the year to about 31,000, after factories cut roughly 113,000 jobs in 2025.
The Federal Reserve also reported that U.S. manufacturing output increased 0.2% in July, reaching its highest level in more than four years. Production of information-processing equipment rose 1.5%, industrial supplies increased 1.4%, semiconductor output climbed 2.4%, and computer and peripheral equipment production rose 1.8%.
However, the recovery is not evenly distributed.
Manufacturing surveys still show weakness across parts of the industrial economy, while sectors connected to AI, semiconductors and infrastructure are benefiting disproportionately. In other words, America’s factory rebound is being powered in part by a highly concentrated investment cycle rather than a universal manufacturing revival.
The Data Center Boom Is Also Reshaping Global Trade
The ripple effect is extending beyond U.S. factories and into international shipping.
The Port of Los Angeles handled 960,464 twenty-foot equivalent units (TEUs) in July, its second-highest July volume on record. Port officials said the flow was supported not only by consumer goods but also by equipment connected to manufacturing and data-center construction.
The changing cargo mix is particularly significant.
Maersk CEO Vincent Clerc said shipping containers are increasingly carrying products linked to electrification and efforts to expand power capacity, including batteries, solar and wind components, turbines and data-center cooling equipment.
That suggests the AI infrastructure boom is beginning to alter what moves through global supply chains — shifting some demand away from traditional consumer merchandise toward equipment needed to build the physical infrastructure behind AI.
But America Has a Major Bottleneck: Power Equipment
The biggest challenge may not be demand. It may be the ability to physically build and power enough data centers.
Transformers, switchgear and other electrical equipment are already facing supply constraints.
Reuters reported in July that lead times for some high-voltage transformers had stretched to multiple years, compared with roughly a year in 2020 and 2021. Data-center construction is also increasing demand for circuit breakers and switchgear.
Wood Mackenzie estimates that U.S. data-center capacity could rise to around 110 gigawatts by 2030, from approximately 24 GW currently. Under an accelerated scenario, data centers could account for as much as 40% of the U.S. electrical-equipment market, compared with just under 2% in 2020.
That creates a paradox: the more aggressively companies build AI infrastructure, the more pressure they put on the equipment needed to connect that infrastructure to the power grid.
Supply Chains Could Become the Next AI Constraint
The pressure extends beyond transformers.
A May analysis from Rabobank found that U.S. data-center expansion is colliding with constraints involving power, water, critical minerals, skilled labor and other supply-chain inputs. It warned that shortages could extend project timelines and make access to physical resources increasingly important when developers choose where to build.
Copper is one example. Data centers depend heavily on copper for electrical systems and other infrastructure, while the same material is also needed for broader electrification. Rabobank noted that copper prices reached record levels in January 2026 amid supply disruptions and strong demand.
The result is a race not simply to secure land and chips, but to secure the physical ingredients required to turn AI investment into operational computing capacity.
The $66 Billion Question
The scale of the opportunity is becoming difficult for manufacturers to ignore.
Wood Mackenzie projects that the U.S. electrical-equipment market tied to data centers could double from $33 billion in 2025 to $66 billion by 2030.
Companies are responding with new factories, expanded production and long-term customer agreements.
Siemens, for example, announced plans to invest more than $200 million in two new plants in Georgia and Texas to produce electrical equipment for data centers and other industrial customers.
Yet manufacturers are also aware of the danger of moving too aggressively.
If AI infrastructure spending slows, companies that have dramatically expanded factories to meet today’s demand could be left with excess capacity. Some manufacturers are therefore using multi-year contracts and other safeguards to reduce the risk of a sudden downturn.
The Bigger Picture
The U.S. AI boom is becoming an industrial story as much as a technology story.
Every new AI data center creates a chain of demand stretching from hyperscale computing companies to generator manufacturers, electrical-equipment producers, construction firms, steel suppliers, cooling companies, logistics providers and raw-material producers.
But the same supply chain that is benefiting from the boom could ultimately determine how quickly the boom can continue.
The race to build AI infrastructure is no longer simply about who has the best chips or the biggest data centers. It is increasingly about who can secure enough electricity, transformers, materials, equipment and factory capacity to actually make those data centers work.
And that may become the next major test for America’s AI ambitions.
Leave a Reply