Bloom Energy surged as much as 15% Tuesday after investors received another signal that the artificial-intelligence power boom may be moving faster than expected. The immediate catalyst looked almost mundane: Bloom acquired a roughly 158,000-square-foot manufacturing facility in Fremont, California. But Wall Street interpreted the expansion very differently. Morgan Stanley analyst David Arcaro said the additional capacity points to strong customer demand, while a separate development from utility Ameren Missouri opened the possibility that fuel cells could expand far beyond individual data centers and into utility-scale power generation. Bloom Energy stock climbed to a three-month high as investors connected those dots.
The rally comes after an extraordinary change in Bloom’s financial profile. Second-quarter revenue reached a record $1.065 billion, up 166% year over year, while product revenue jumped 215%. Gross margin improved to 33.4%, operating income reached $182.2 million and operating cash flow swung to positive $226.4 million. Management responded by raising full-year 2026 revenue guidance to between $3.9 billion and $4.2 billion, representing roughly 100% year-over-year growth at the midpoint.
Those numbers explain why a factory purchase can suddenly move Bloom Energy stock by double digits. Investors are no longer asking whether fuel cells can find a market. They are asking whether Bloom can manufacture enough of them to satisfy a market that may be expanding from AI data centers into utilities, industrial facilities and other power-hungry infrastructure.
That is a dramatically different problem — and potentially a much more valuable one.
New Factory Is Really a Bet That the Orders Keep Coming
The Fremont expansion gives Bloom additional manufacturing space at a moment when power availability has become one of the AI industry’s biggest constraints.
Related articles
Bloom’s solid-oxide fuel cells generate electricity onsite, allowing customers to build power infrastructure without waiting entirely on conventional utility-grid connections. That has become increasingly valuable because data-center developers can face years-long delays for transmission upgrades, transformers and new generating capacity.
The broader market is scrambling for alternatives. Enverus Intelligence Research estimates that nearly 30 gigawatts of behind-the-meter gas generation could be added in the United States by 2030, with data centers accounting for approximately 88% of that capacity. Demand has become so intense that developers are increasingly considering smaller turbines and other onsite technologies despite potentially higher lifetime power costs, simply because electricity is needed faster than traditional infrastructure can deliver it.
Unlike conventional gas turbines, its fuel-cell systems generate electricity electrochemically rather than through combustion. The company argues this allows systems to be installed quickly, operate continuously and serve power-intensive facilities directly at the site.
Management says the technology has already passed a critical commercial threshold. During Bloom’s second-quarter results, CEO KR Sridhar said all major U.S. hyperscalers, along with more than a dozen neoclouds, AI laboratories and colocation data-center operators, had validated and approved Bloom’s power solutions for AI facilities.
That makes the Fremont expansion more than another factory.
Bloom is effectively betting that AI’s electricity shortage will last long enough to justify building considerably more manufacturing capacity.
The Numbers Behind Bloom Energy Stock Have Changed Astonishingly Fast
Only a year ago, investors could reasonably question whether Bloom’s growth story would ever translate into consistent profitability.
The company’s recent results have made that argument considerably harder.
First-quarter 2026 revenue reached $751.1 million, compared with $326 million a year earlier. Non-GAAP gross margin climbed to 31.5%, while adjusted EBITDA reached approximately $143 million. Bloom initially raised its full-year revenue outlook to between $3.4 billion and $3.8 billion.
Then the second quarter changed the trajectory again.
Revenue accelerated to $1.065 billion, crossing the $1 billion quarterly threshold for the first time. GAAP gross margin rose above 33%, operating income reached $182.2 million and EPS climbed to $0.62 from a loss of $0.18 in the prior-year period. Bloom subsequently raised full-year revenue guidance again, this time to $3.9 billion to $4.2 billion.
The acceleration is important because manufacturing expansion becomes much easier to justify when factories are supporting proven revenue growth rather than merely anticipated demand.
Bloom’s existing AI infrastructure business also continues broadening. In August, the company expanded its partnership with MiTAC Computing Technology, agreeing to deploy an islanded fuel-cell microgrid at MiTAC’s AI-server manufacturing campus in Fremont. Bloom said its AI infrastructure segment at that point spanned nearly two dozen customers representing approximately 250 megawatts of capacity.
Oracle’s 2.45-Gigawatt Project Shows How Big Bloom Could Become
Oracle’s Project Jupiter in New Mexico represents the clearest demonstration of what the next phase of Bloom’s growth could look like.
In April, Oracle announced that Project Jupiter would use Bloom fuel cells to fully power the massive AI data-center campus, with up to 2.45 gigawatts of installed Bloom capacity. The revised design replaced previously planned gas turbines and diesel generators with a fuel-cell microgrid.
To put that number in perspective, Project Jupiter alone represents almost ten times the approximately 250 megawatts of AI infrastructure capacity Bloom referenced across nearly two dozen customers in August.
That scale explains why investors reacted nervously when Oracle issued a force majeure notice connected with potential delays at the New Mexico project last week.
The problem centers partly on the natural-gas infrastructure required to support the facility. Regulatory and infrastructure complications have created uncertainty around timing, although Oracle has maintained that Project Jupiter remains on schedule. Reuters reported that Oracle’s notice was designed to protect it financially if the facility fails to meet its targeted 2028 operational timeline rather than terminate its commitment to the project.
Bloom shares initially came under pressure as investors worried that one of the company’s most important future projects could be endangered.
Morgan Stanley reiterated that the issues were not related to Bloom’s technology, while Oracle and Bloom continued to indicate that the project was moving forward. Morgan Stanley’s Arcaro also noted that Project Jupiter was not incorporated into Bloom’s 2026 guidance, reducing the risk that delays would undermine this year’s financial outlook.
A 500-Megawatt Utility Plan Could Be Even More Important Than the Factory
The second catalyst behind Tuesday’s rally received less attention than the Fremont expansion but could ultimately prove more important.
Ameren Missouri released its 2026 Integrated Resource Plan on Monday, outlining how the utility expects to meet rapidly growing electricity demand over the next two decades. Among the proposed resources is 500 megawatts of natural-gas fuel cells by 2030.
Ameren described fuel cells as a promising, scalable technology capable of delivering reliable electricity at any time of day.
Bloom has not been announced as the supplier for those 500 megawatts, so investors should not treat the plan as a Bloom contract.
But Wall Street immediately recognized the potential implication.
Bloom’s recent growth has been driven largely by customers that need onsite electricity because conventional grid infrastructure cannot arrive quickly enough. If regulated utilities themselves begin incorporating hundreds of megawatts of fuel cells into long-term generation plans, Bloom’s addressable market could expand dramatically.
Ameren’s plan also reveals why utilities are searching for new solutions. The company says it has signed electric-service agreements covering 2.8 gigawatts of aggregate large-load demand by 2030, with additional growth possible. Its preferred plan includes enormous additions of natural gas generation, batteries, solar, wind and eventually nuclear power alongside the proposed fuel-cell capacity.
This is no longer merely a data-center story.
AI-driven electricity demand is beginning to reshape how utilities plan their entire generation portfolios.
If fuel cells secure a meaningful role in that transition, Bloom could find itself competing for a market far larger than hyperscaler campuses alone.
The AI Power Shortage Is Becoming Competitive Advantage
Traditionally, large electricity users could connect to the grid and rely on utilities to provide additional generating capacity. AI has broken that model in many locations because demand is arriving much faster than new transmission lines and conventional power plants can be built.
Developers increasingly care about one question: how quickly can electricity reach the GPUs?
That makes technologies capable of being deployed directly at a data center more valuable.
Bloom is pushing the advantage further with its 800-volt direct-current architecture. The company says next-generation AI computing infrastructure will increasingly use 800V DC internally, while conventional grid power arrives as alternating current and requires multiple conversion stages.
Bloom argues that its fuel cells can produce 800V DC power directly. According to a company analysis released September 16, a 1-gigawatt AI facility using its DC-native architecture could reduce non-compute capital expenditures by approximately $3.6 billion, or 27%, and lower five-year total cost of ownership by $5.5 billion compared with traditional AC configurations. Those are Bloom’s estimates rather than independently established industry figures, but they illustrate the economic argument the company is making to hyperscalers.
If customers validate those economics at scale, Bloom’s competitive position could become considerably stronger.
But investors also need to understand what today’s stock price is already assuming.
Bloom Energy Stock Is No Longer Being Valued Like a Niche Fuel-Cell Company
The spectacular performance of Bloom Energy stock reflects a major reclassification taking place in investors‘ minds.
Bloom was once viewed primarily as a clean-energy or hydrogen-related company. Today, the market increasingly treats it as an AI infrastructure company.
That distinction matters because AI infrastructure businesses have commanded much richer valuations as investors search for companies positioned around the physical bottlenecks preventing data-center expansion.
Power is arguably becoming the biggest bottleneck of all.
Morgan Stanley’s Arcaro maintained an Overweight rating and a $310 price target following concerns around Project Jupiter. Tuesday’s factory expansion subsequently reinforced his view that Bloom is experiencing strong customer demand.
Yet expectations have risen alongside the share price.
That creates a different kind of risk. Bloom no longer needs simply to prove that its fuel cells work or that customers want them. It must demonstrate that it can manufacture systems rapidly, maintain margins as volumes increase and execute enormous deployments without project delays destroying the economics.
Project Jupiter illustrates that problem perfectly.
Bloom’s technology can be ready while pipelines, permits or other infrastructure are not. The company can therefore face project-timing risk even when the underlying demand remains intact.
The larger Bloom becomes, the more those execution details will matter.
The Next Test Is Whether Bloom Can Turn Power Scarcity Into Durable Profits
Tuesday’s rally was triggered by a factory expansion, but the market is really reacting to something much larger.
Bloom Energy is positioning itself at the intersection of two extraordinary trends: the AI industry’s seemingly insatiable demand for electricity and America’s inability to build conventional power infrastructure quickly enough to satisfy it.
The company’s financial results suggest that opportunity has already moved beyond theory. Revenue reached $1.065 billion last quarter, year-over-year growth hit 166%, operating cash flow turned strongly positive and management doubled the midpoint of its 2026 revenue outlook compared with the previous year’s revenue base.
Now the opportunity is expanding again.
The Fremont facility suggests Bloom expects more manufacturing capacity to be needed. Oracle’s Project Jupiter demonstrates that individual AI campuses can require gigawatts of onsite power. And Ameren Missouri’s plan to add 500 megawatts of natural-gas fuel cells by 2030 suggests the technology could eventually move from corporate microgrids into mainstream utility generation.
None of that guarantees Bloom will capture every opportunity. Ameren has not selected Bloom for its proposed fuel-cell capacity. Project Jupiter still faces infrastructure and regulatory complications. And a stock priced around enormous future growth becomes increasingly vulnerable whenever execution fails to match expectations.
But Bloom’s newest factory sends an unmistakable message.
Management appears to believe the problem ahead is no longer finding customers for fuel cells. It is building enough fuel cells quickly enough to satisfy them.
If that assessment proves correct, Bloom’s transformation from niche clean-energy company into one of the AI boom’s critical power suppliers may still have another chapter to run.
The next question for Bloom Energy stock is whether the factories can expand as quickly as the demand Wall Street now expects them to serve.
Disclaimer
This article is for informational purposes only and does not constitute financial or investment advice. Readers should conduct their own research or consult a qualified financial advisor before making investment decisions. This article was researched and drafted with the support of AI, but was reviewed, fact-checked, and edited by the editorial team before publication.
