Bloom Energy Stock: The Hottest Name in the AI Power Trade

Bloom Energy stock is one of the most searched names in the AI energy theme, and the search volume makes sense. Of all the companies attacking the data-center power shortage, Bloom is the purest publicly traded bet on on-site generation: solid-oxide fuel cells that convert natural gas or hydrogen into electricity at the point of use, with no grid connection required. When AI operators started hitting grid interconnection queues measured in years, that profile went from niche to necessary.

Jon Najarian’s Infinite Power presentation for Banyan Hill is the latest entry in the AI energy pitch wave, and while his actual pick sits behind the paywall, the theme he builds it on is the same one driving Bloom’s story. Our full teardown covers the presentation. This piece explains the Bloom thesis on its own merits.

What Bloom Energy Does

Bloom’s “Bloom Energy Server” boxes generate electricity through electrochemical conversion rather than combustion. They are modular, they can run on natural gas today and hydrogen tomorrow, and they can be sited directly at a data center, which is the entire point in a market where the local utility cannot deliver. That is the behind-the-meter pattern: generation that skips the queue entirely.

The documented context around the theme is stark. AI data-center power demand is outrunning grid capacity, with BlackRock’s cited 148-gigawatt figure reflecting genuine published estimates, and the Department of Energy warning that blackouts could increase a hundredfold by 2030 under certain conditions. Najarian’s Memphis case study is the emblem of the workaround: xAI needed 150 megawatts, the utility could supply about 8, so the company built its own on-site plant with truck-delivered gas turbines in about 122 days.

Why the Market Watches Bloom

Bloom is the name most investors find first when they go looking for that story in public markets, and its major-tech partnership announcements and utility-scale fuel cell deals have made it the sector’s momentum name. The bull case is straightforward: if on-site generation becomes standard practice for hyperscalers, Bloom’s order book compounds. The bear case is equally straightforward: fuel cells compete against gas turbines and diesel gensets on cost per megawatt, and Bloom has a long history of thin margins to overcome.

That is the honest frame for any AI energy holding: the demand story is real and documented, and the competition for the supply contract is fierce.

Where Bloom Sits Relative to the Pitch

Najarian’s teased company is a gas-turbine lessor, not a fuel-cell maker, so Bloom is not the pick; the clue profile in the teardown, a Houston company leasing power pods under 10-to-15-year contracts with xAI as anchor customer, points elsewhere. But the structural insight the pitch sells, “power as a service,” applies to the whole sector: the companies that lease generation rather than sell it once collect recurring revenue through the contract term, which changes how the market values them. Bloom’s own business, selling boxes plus service contracts, sits adjacent to that shift.

The pattern stocks Najarian cites to frame the opportunity are the real history lesson: Vertiv’s 1,200% run as the data-center power supplier, Generac’s 850% pandemic surge, and Quanta Services’ 8,600% multi-decade climb as the grid contractor. Whether Bloom follows that trajectory is the open question every AI energy investor is asking, and the honest answer is that the demand side supports the thesis while the margin side keeps it contested.

Bloom’s Own Data Inside the Pitch

One detail worth knowing when evaluating this presentation: the most striking behind-the-meter statistic in it comes from Bloom Energy’s own research. The claim that “more than 60 percent of new data center developers are now planning to build their own generation” is attributed to Bloom’s data center power report, and it checks out against published coverage. That is a candid source in one sense, since Bloom sells exactly the equipment that trend would favor, and the presenter quotes it accurately rather than dressing it as independent analysis.

The same honesty test applies to the pattern stocks. Vertiv’s 1,200% five-year run, Generac’s 850% surge, and Quanta Services’ 8,600% lifetime return are all real price history, and the presentation presents them as what the supplier pattern has done before, not as promises about the current pick. Bloom sits in the middle of that family: a real company with real fuel-cell deployments and a documented push into AI data center power, but also a stock whose story rises and falls with hyperscaler capex cycles. Anyone weighing the theme should hold both facts at once. The AI power shortage is quantified and real, the IEA’s projection that data center electricity use roughly doubles by 2030 reflects genuine published analysis, and Bloom is one credible way the market is solving it, without being the only one.

What to Watch

For readers evaluating the AI power theme, the verifiable signposts are the grid statistics from the pitch, which check out against the ASCE’s 2025 report: 70% of power transformers are 25 years or older, 60% of circuit breakers are 30 years or older, and the grid’s D+ grade is real. Every year of grid delay pushes more demand toward on-site generation, whichever technology wins it.

For the demand story from other angles, our data center power demand piece and the Homestacks coverage are good companions, and the Musk AI energy bottleneck piece covers the same shortage thesis from a different publisher’s framing.

The short version on Bloom Energy stock: real company, real technology, sitting in front of a real and quantified power shortage. The contest is on cost and margins, and that contest is why the stock trades with more volatility than its theme deserves credit for.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.