Top 3 Electricity Stocks to Watch in the AI-Data-Center Boom

AI is booming. We've seen it in the rise of semiconductor stocks. For evidence, look no further than the iShares Semiconductor Fund (SOXX). The exchange-traded fund ("ETF”) holds several familiar semiconductor names – Nvidia (NVDA), Advanced Micro Devices (AMD), Broadcom (AVGO), and many more. Year to date, SOXX is up close to 40%, more than double the performance of the S&P 500 Index.

But the AI build-out is having an impact on other sectors not directly related to technology. We covered how building materials and construction companies are benefiting from AI data centers. We've also examined how natural gas serves as a near-term energy solution to power data centers... and how nuclear fusion may be the long-term answer.

But another area investors should keep on their radars is electrification and thermal management. In plain English, those would be companies that keep electricity flowing to the right places in data centers and keep the hardware cool.

These companies are already benefitting from the data-center build-out. And the reason is straightforward.

AI's Soaring Electricity Demand Is Reshaping the Energy Market

The concept is simple: AI data centers need energy, and plenty of it. As Nick Koziol shared recently, the new power demand estimates are staggering.

Here's Nick sharing the recent data:

In the U.S. alone, data centers accounted for 4.4% of total electricity usage in 2023.

But with the growth of AI, that share is going to grow significantly. The Lawrence Berkeley National Laboratory estimates that, by the end of 2028, data centers will account for 6.7% of total electricity demand – at the very least.

In their "bull" case, data centers will use as much as 12% of all electricity consumed in the U.S. So, in just five years, AI will force data centers to grow their electricity usage by 50% (at the low end) or by 170% (at the high end).

These facilities also generate extreme heat. Graphics processing unit ("GPU”) packages operating under heavy computational loads can reach temperatures of 105 degrees Celsius. That would be 221 degrees Fahrenheit, more than hot enough to fry an egg.

And that means that liquid cooling is vital. That's when a liquid coolant absorbs and dissipates heat from IT equipment, primarily the AI accelerators within servers. These include GPUs, Tensor Processing Units ("TPUs"), and application-specific integrated circuits ("ASICs").

Power distribution providers are essential for optimal and efficient AI-data-center performance. As mentioned above, these companies help manage the astounding power demands of AI workloads and keep it running cool in the process.

Let's have a look at some of the companies that help do just that...

Power Infrastructure Leaders Poised to Benefit From AI Growth

Nearly every component of an AI data center requires electricity. Power distribution is the system that delivers electricity to servers. Redundancy is the design strategy that uses multiple, independent power paths to prevent downtime during a power failure. AI data centers require this redundancy because AI workloads need nonstop power.

And because those workloads spike at times, power loads become unpredictable. For instance, back in March, a new trend took hold among ChatGPT users who were creating images of themselves in the style of Studio Ghibli, a Japanese animator. But the fad caused a huge spike in computational demand. That meant a huge spike in power consumption and heat generation.

Source: X.com/@sama

Companies like Ohio-based Vertiv (VRT) specialize in the critical power and integrated infrastructure that supports AI-data-center workloads. Some of Vertiv's core offerings include:

• AI-ready uninterruptible power supply ("UPS") systems that can effectively handle rapid power load changes, ensuring smooth and consistent operation.

• Power distribution units ("PDUs") designed for the high-voltage needs of AI centers.

• Battery energy storage systems ("BESS") and other solutions that allow data centers to build microgrids, which reduces dependency on the utility grid.

Another company to watch is nVent Electric (NVT), a London-based power management company. nVent's high-density, intelligent power distribution units ("IPDUs") help manage electricity for AI workloads.

One more is Eaton (ETN). Like the two mentioned above, the company manages the power demands of data centers. Eaton has also collaborated with Nvidia on an 800-volt direct current power architecture that improves power efficiency while supporting the high demands of AI workloads. Eaton also developed what's called "busbar technology," which replaces traditional cables with rigid metal conductors that are more compact and reduce power loss. And like Vertiv and nVent, Eaton also produces high-density PDUs to handle the power needs of AI racks.

It's not just power management. All three also provide services for cooling, primarily liquid cooling. They also offer software to their clients to track performance, predict failures, and increase efficiency within data centers.

Put simply, when it comes to managing power in data centers, keeping it all cool, and overseeing the entire operation, these are three of the names we're keeping an eye on. And we're not the only ones. On a year-to-date basis, NVT, VRT, and ETN are up 60%, 52%, and 12%, respectively.

Eaton's relative underperformance isn't too surprising. As a larger, more established company, the stock's fundamentals are growing thanks to the data-center build-out, but just not as fast as the other two. Looking at the latest quarterly revenue figures for each, we see:

• Eaton: Approximately $7 billion, 10.6% year-over-year growth

• nVent: Approximately $963 million, 30% year-over-year growth

• Vertiv: Approximately $2.6 billion, 29% year-over-year growth

Why Electrical Stocks Deserve Investor Attention

The AI explosion, financially speaking, is coming in waves. The first wave, driven by Nvidia, saw investors – and the federal government – going all in on semiconductor and chip stocks. Consider these statistics:

• In 2025, the Morningstar Global Semiconductors Index was up 34% through September... more than twice the return of the overall U.S. market.

• NVDA stock rose almost 270% between early 2024 and early October 2025.

• Taiwan Semiconductor Manufacturing (TSM) stock exploded at near-Nvidia levels, jumping 250% between January 2023 and October 2025. In Q3 2025 alone, the company's revenue grew 41% primarily due to AI chip demand.

• In 2024, AI chips made up around 20% of the industry's total revenue, even though their volume was only 0.2% of all manufactured.

• The global AI chip market is projected to hit $332.77 billion by 2030.

These semiconductors may be the workhorse for AI computing, but they're not the only piece of the puzzle. As we just showed, power management products are playing a key role in keeping data-center operations running smoothly.

And as for whether the names listed above are worthy of consideration for investors, we turn to our proprietary Stansberry Score. First up is Vertiv.

Vertiv gets a respectable overall grade of B. That score is driven higher by Vertiv's financials, which are ranked in the top 200 of the 5,000-plus companies our Stansberry Score grades. But the company gets dinged on "Capital Efficiency" and "Valuation." The first one isn't much of a surprise. The types of products and services Vertiv provides are capital-intensive.

As for valuation...

Vertiv is expected to report earnings per share ("EPS") of $4.08 by the end of 2025. With a share price in the ballpark of $190, that implies that VRT is trading at 46 times full-year earnings. That's about twice as expensive as the S&P 500, which is around 23 times forward earnings.

Vertiv is expected to grow quickly: 2026 EPS is expected to come in around $5.14, or 25% year-over-year growth. But, make no mistake, investors are paying up for that growth.

Moving on to nVent, we see a similar score of B. nVent gets high marks in capital efficiency, showing a business that can generate real returns for investors. But, like Vertiv, it gets dinged only slightly on valuation... and for the same reasons.

As for Eaton, it scores the highest of the bunch, with an overall grade of A.

Eaton's financials are ranked very high – 47 out of the 5,000-plus stocks covered. Capital efficiency is also above average. It's only on valuation that we see a lower grade for Eaton. But, once again, this is a quality business. Eaton went public in 1968 and has been printing steady returns for literally decades. And great companies like that usually garner a higher price tag.

This is the market recognizing how vital electrical services, infrastructure, and management are to AI data centers. And this is a sector investors should pay close attention to.

That said, the same trend of increased AI-data-center spending that has lifted these stocks would also become a headwind if that spending ever slowed. Put another way, watch out if the AI boom turns into a bust.

(My colleague Alan Gula wrote an excellent piece detailing why he thinks the AI boom still has some room to run.)

The Bottom Line

It's difficult to predict whether we'll reach the point of an "AI bubble." Right now, AI is booming. But investors don't need to buy Nvidia, Oracle (ORCL), or Taiwan Semiconductor to gain exposure to this trend.

Investors can find wins in the industries that are building the AI infrastructure and powering the data centers. That includes electricity, building materials, and AI cloud services – the next wave of the AI boom.

Regards,

David Engle

Editor’s Note: The AI trade is evolving – and the biggest winners may not be the chipmakers, but the electricity stocks keeping AI data centers online. To know when to buy, hold, or sell these plays, we recommend Whitney Tilson's N.E.W. Wealth System – a proven framework for profiting from megatrends like AI, infrastructure, and energy. Click here to learn how Tilson's system helps investors capture the next phase of the AI trade.