The artificial intelligence boom’s forgotten winners are hiding in plain sight

Key takeaways

• Artificial intelligence needs cooling, electricity, construction and automation before it can produce a useful answer.

• Industrial suppliers may benefit from wider demand, but capacity, execution and valuation still matter.

• Investors can study the full data-centre chain instead of treating artificial intelligence as one crowded technology trade.

Everyone talks about artificial intelligence models, but the boom is also making factories exciting again. Yet far less attention goes to the companies supplying the picks and shovels that make it all possible.

A modern data centre is a specialised factory. It consumes enormous amounts of electricity, produces serious heat and cannot afford to stop. The International Energy Agency expects global data-centre electricity use to more than double by 2030. That means more cables, transformers, cooling systems, backup power and construction work.

The boom is widening from software and semiconductors into the physical economy.

The chatbot needs a boiler room

The first layer sits inside the data centre. Vertiv provides power management and cooling equipment. Trane Technologies supplies large cooling systems, including liquid cooling, which carries heat away from powerful chips.

Schneider Electric and Eaton sit at the electrical heart of the building. They provide switchgear, power controls and backup systems that safely move electricity to thousands of servers. Legrand adds smaller but essential components, including power distribution systems, racks and connections.

These businesses benefit from a simple truth: more computing creates more heat and requires more electricity. A better model may be digital, but the supporting infrastructure remains stubbornly physical. Servers still dislike heat, power cuts and improvisation.

Recent results support the theme. Schneider reports strong growth led by data centres. Eaton says data-centre orders in its main United States electrical business have risen sharply, while Vertiv continues to report rapid sales growth. The market has noticed, so the opportunity is less hidden than the headline suggests.

Artificial intelligence meets the grid

The second layer begins outside the building. Data centres can require as much power as a small industrial site, but electricity grids were not designed for demand to arrive this quickly.

Siemens Energy and GE Vernova sell turbines, transformers, switchgear and grid equipment. Their products help generate electricity and move it to where it is needed. Both are seeing strong demand linked partly to data-centre development.

ABB offers electrification and automation equipment across factories, buildings and utilities. Its role shows why the theme reaches beyond artificial intelligence. The same equipment also supports grid upgrades, factory expansion and wider electrification.

Artificial intelligence is one driver, while ageing grids, energy security and manufacturing investment add others. That mix may make demand more durable, although it also makes these companies harder to value with one simple label.

The builders behind the builders

The final layer is execution. Quanta Services builds and upgrades power networks. Comfort Systems installs electrical, heating and cooling systems in complex buildings, including data centres.

These contractors benefit when spending moves from announcements to construction. They also reveal the bottleneck. Money alone does not build a data centre. Projects need skilled workers, permits, transformers and grid connections. Long waiting times may support supplier pricing, but they can also delay revenue and increase costs.