Summary:
To ring in 2026, our 24/7 Wall St. Analysts Eric Bleeker and Austin Smith are counting down 12 trends for AI investors to watch in 2026.
The second trend the two have discussed is the rise of AI factories – a new generation of massive data centers designed to manufacture intelligence at industrial scale. These facilities, often exceeding one gigawatt in size, represent a dramatic step change from traditional hyperscale data centers and signal a fundamental shift in how AI infrastructure is built.
The rapid expansion of AI factories is creating entirely new demands across electricity delivery, construction, networking, and cooling. As computing density rises, thermal management has become a critical constraint, accelerating the transition from legacy air cooling to advanced liquid cooling solutions. This shift creates a c
“This is an area that Jensen Huang and other leaders in the space have been talking about,” Bleeker explains. “Microsoft has even said it now sees itself as a modern construction company, and what it’s building are AI factories.”
As Bleeker continue, AI-driven infrastructure growth is reshaping multiple industries and creating a new class of long-duration investment opportunities.
“When you get a step change like this, you want to find areas where demand is exploding at the same time legacy technology is being replaced. This new class of AI factories has entirely new demands: how electricity is delivered, who builds them, how they’re cooled, and how networking is handled.”
If the video embed above isn’t working, you can watch the original video at: https://www.youtube.com/watch?v=76aig4X3dm4
Don’t Ever Miss an Episode of the AI Investor Podcast
If you’re looking for AI stocks that could have major catalysts in 2026, make sure to check out The AI Investor Podcast. It’s free to subscribe in your favorite podcast player and every episode we give updates and recommendations in a $500,000 portfolio filled with our top AI stock ideas.
You can listen to our most recent episode in either Apple Podcasts below:
Transcript:
Austin: Let’s look at your next trend: AI factories. Is this about humanoids or purpose-built robots? One thing you said earlier that really stuck with me is that Amazon (NASDAQ: AMZN) | AMZN Price Prediction may be one of the biggest beneficiaries of the AI robotics revolution. Amazon already heavily uses single-purpose robots, and we discussed data last year showing that Amazon now has more robots handling packages than humans. To the extent Amazon is showing what’s possible, I imagine that’s what underlies this trend—but what you’re really talking about is AI factories for everyone.
Eric Bleeker: This is an area that Jensen Huang and other leaders in the space have been talking about. Microsoft (NASDAQ: MSFT) has even said it now sees itself as a modern construction company, and what it’s building are AI factories.
There’s no formal definition, but I roughly define AI factories as data centers larger than one gigawatt. For perspective, when we started this podcast, we were amazed by Elon Musk pushing xAI to build a 100-megawatt facility. The Stargate project I visited in Abilene is now 1.9 gigawatts – nineteen times larger in a single project.
To clarify, these are data centers, but they are manufacturing intelligence at industrial scale. I live in Northern Virginia’s “Data Center Alley,” which hosts about 35% of hyperscalers – generally 50 megawatts or larger. What we’re now talking about is a minimum of 20× that size, and in many cases 100× larger.
Wood Mackenzie is tracking roughly 245 gigawatts of data center capacity across the U.S. Texas’s ERCOT grid alone has 225 gigawatts in some form of development, compared to just seven gigawatts last year. This is a massive step change.
When you get a step change like this, you want to find areas where demand is exploding at the same time legacy technology is being replaced.
Austin: Right – there’s a compounding effect when demand surges just as a new technology replaces the old one.
Eric Bleeker: Exactly. We’re seeing that now in data centers. This new class of AI factories has entirely new demands: how electricity is delivered, who builds them, how they’re cooled, and how networking is handled.
At CES, NVIDIA (NASDAQ: NVDA) made headlines when Jensen Huang suggested their next-generation chips might not require liquid cooling. The nuance is important. Liquid cooling is still projected to grow about 15× by the end of 2027, while traditional air cooling shrinks.
Liquid cooling includes cold plates on chips, coolant distribution units on racks, and large-scale chillers. The reality is that all this heat must be dissipated. The market panicked on headlines, but many companies sold off due to misunderstandings rather than fundamentals.
Austin: And Alphabet (NASDAQ: GOOGL)’s TPU chips are designed specifically for liquid cooling. Those chips were a major reason the stock performed so well at the end of 2025. If TPUs become more prominent because of their efficiency, that implies strong demand for liquid cooling.
Eric Bleeker: Exactly. Liquid cooling has gone from extremely niche last year to broad adoption. You get a double benefit: rising data center demand overall and rising adoption of this specific technology.
We’ve recommended Vertiv (NYSE: VRT) and Schneider Electric (OTC: SBGSF) – large, diversified players in this space. We also like Modine Manufacturing (NYSE: MOD), which focuses on thermal management. It’s much smaller, but earnings growth has been exceptional and estimates suggest they could more than double earnings over the next three years.
Another interesting company is nVent Electric (NYSE: NVT). It has two growth engines: data centers – particularly liquid cooling – and power infrastructure. About 40% of its sales now come from AI-related demand, with accelerating growth rates.
Many of these stocks have sold off due to headline-driven fear, even as demand trends continue to improve. With gigawatt-scale AI factories still largely ahead of us, there may be a “Goldilocks” period where fundamentals improve faster than market expectations.
Austin: We all love Goldilocks moments here on the AI Investor Podcast.
Top Gaining Stocks
Top Losing Stocks
