The AI Data-Center Boom Is Creating a New Class of U.S. Manufacturing Winners

The artificial-intelligence boom is increasingly an industrial story. Every large data center needs dependable power, electrical distribution, cooling, mechanical systems and construction materials, creating new demand for manufacturers that may have had little connection to the technology sector. 20

Generac provides the clearest example. The Wisconsin-based generator maker plans to spend $250 million by the end of 2027 to equip multiple factories to produce larger systems for data centers. Its data-center-generator backlog has reached about $1.6 billion, and the company expects to add roughly 1,000 workers as it expands capacity. 320

Why data centers are pulling manufacturers into the AI economy

Data centers are large, power-intensive construction projects. Their supply chains include:

• Backup generators and other power-generation equipment
• Transformers, switchgear and electrical distribution systems
• Cooling and exhaust systems
• Construction machinery and industrial bearings
• Electric cable, pipe, cement and steel components
• Hoses and prefabricated metal building materials 1020

That breadth matters. The AI buildout is not benefiting only chip designers, cloud companies and server manufacturers. It is also sending orders to electrical-equipment makers, regional fabricators and established industrial suppliers that provide the physical systems needed to keep computing facilities operating.

Generac shows how the boom is changing industrial demand

Generac is best known for residential backup generators, but data-center demand is now strengthening its commercial and industrial business. In the second quarter of 2026, the company reported $1.17 billion in net sales, while commercial and industrial sales rose about 29% year over year to $556 million. Data-center shipments helped offset weaker conditions in some residential and portable-generator categories. 4

The shift also changes the company’s operating challenge. Generac must add factory capacity, hire workers and deliver large systems on schedules set by hyperscale customers. Its data-center opportunity is substantial, but it is more concentrated in a single fast-growing market than its traditional consumer business.

The opportunity reaches smaller industrial suppliers

The effect is visible among specialized manufacturers as well. Georgia-based Southeastern Hose, which traditionally served steel and petrochemical customers, is supplying corrugated metal hoses used in data-center cooling and exhaust applications. Its management described orders that moved from an initial trickle to a major surge as data-center construction accelerated.

This is an important feature of the boom: suppliers do not necessarily need to sell software, chips or servers to benefit from AI spending. A company with the right manufacturing capability, certification or industrial component can become part of the data-center supply chain when a new application emerges.

A rapidly expanding market for electrical equipment

Wood Mackenzie projects that U.S. data-center electrical-equipment spending will rise from $33 billion in 2025 to $66 billion in 2030. 20 That forecast helps explain why manufacturers are investing in plants and production lines even as parts of the wider manufacturing economy remain under pressure.

Siemens has also reported strong demand connected to AI data centers, alongside demand from electronics and semiconductor manufacturers expanding capacity for AI-related equipment. The company reported its highest-ever quarterly industrial profit and raised its full-year guidance in August 2026. 19

The result is a two-speed industrial economy: companies exposed to data-center power and construction can be expanding factories and order books while manufacturers tied more closely to consumer demand face softer conditions.

More factory orders do not automatically mean broad wage growth

The employment impact is also more complicated than the headline hiring announcements suggest. Generac’s planned 1,000 additions are a direct example of new manufacturing demand, but local effects vary by industry and location.

Research from Brookings found that U.S. labor markets receiving their first large data center experienced higher employment in data processing and telecommunications over the following decade. However, the study found no wage increase, and estimated that a typical treated county gained roughly 100 to 200 jobs depending on facility type.

In other words, data centers can create meaningful local employment while producing benefits that are concentrated in particular occupations, suppliers and regions rather than lifting every part of the surrounding economy equally.

Capacity shortages could limit the upside

The biggest near-term constraint is not a lack of orders but the ability to fulfill them. U.S. power companies and data-center developers are competing for scarce equipment, particularly transformers. Shortages have pushed up costs, lengthened wait times and encouraged buyers to place orders well in advance.

Manufacturers also face limited factory space, shortages of skilled labor and volatility in the price and availability of imported materials and components. Those pressures can turn strong demand into project delays, higher construction costs and lower margins for suppliers that expand too quickly.

The risk: an AI infrastructure overbuild

The same concentration that makes the data-center market attractive also creates risk. Large technology companies have committed about $1.09 trillion in future lease payments, mostly connected to data centers that have not yet begun operating. 17 Those commitments demonstrate the scale of the buildout, but they do not guarantee that every planned facility will generate adequate returns.

A slowdown could affect the supply chain in several ways. Developers could defer or cancel projects, equipment orders could be pushed back, and manufacturers that added capacity for AI customers could be left with underused factories. Scarce components could also be diverted away from long-standing industrial customers during the boom and then become less valuable if demand falls.

That concern is not hypothetical: a Reuters analysis has argued that the economics of some new AI data centers could prove difficult if revenues and utilization do not match the scale of the investment. The outcome will depend on how quickly AI services generate durable demand, not simply on how many facilities companies announce.

The broader lesson for manufacturers

AI data centers are currently functioning as a powerful new source of U.S. industrial demand. Generac’s backlog and factory expansion show the scale of the opportunity, while the experience of specialized hose suppliers illustrates how far the effect can spread through regional manufacturing networks.

But the strongest winners may be companies that treat data-center demand as an opportunity to diversify rather than as a reason to abandon existing customers. Capacity discipline, long-term contracts and a balanced customer base could matter as much as the size of today’s order book.

The AI buildout is creating real manufacturing winners—from generator makers to cooling-equipment and component suppliers. The open question is whether that demand becomes a durable industrial cycle or an unusually concentrated expansion that leaves excess capacity behind when the economics of AI infrastructure are tested.