Bloom Energy ((BE Quick QuoteBE - Free Report) ) has been a big winner in the AI datacenter buildout because they offer an immediate "BTM" (behind the meter) solution where they can deploy solid oxide fuel cells (SOFCs) in under 90-days for safe, clean, zero-combustion power.

I just profiled the company and its prospects in this video on Monday. We've owned the stock since the $70s after I learned of their partnership with Oracle ((ORCL Quick QuoteORCL - Free Report) ) and continue to enjoy opportunities to swing trade it above $250, all while holding a core position for higher targets in my TAZR Trader portfolio. 

After we bought shares in September, we continued to accumulate as the good news really started to heat up with this October "bloom"...

Brookfield ((BAM Quick QuoteBAM - Free Report) ) and Bloom Energy announced a $5 billion partnership to build global AI infrastructure facilities, with Bloom Energy serving as the preferred onsite power provider for Brookfield’s new AI factories.

The deal marks Brookfield’s first investment under its AI Infrastructure strategy and will use Bloom’s fuel cell technology to power data centers requiring massive, always-on electricity capacity. The companies said a European site will be announced later this year.

“AI infrastructure must be built like a factory, with purpose, speed, and scale,” said Bloom CEO KR Sridhar, adding that today’s grids “cannot support” the real-time load responsiveness AI factories need.

Brookfield’s Sikander Rashid said “behind-the-meter power” is key to closing the grid gap for AI data centers, which could push U.S. AI-related power demand above 100 GW by 2035, according to industry projections cited in the release.

What was so odd to me afterwards was that major Wall Street i-banks like Bank of America and Jefferies maintained "Sell" ratings and sub-$40 price targets throughout Q4.

Then in January, American Electric Power ((AEP Quick QuoteAEP - Free Report) ) announced a $2.65B SOFC deal with Bloom. The major fuel cell procurement agreement was tied to the development of a fuel cell power generation facility in Wyoming, underscoring the scale of the planned buildout.

According to AEP, its unregulated subsidiary entered into an agreement with Bloom Energy in November 2025 to acquire 100 megawatts of solid oxide fuel cells, while also securing an option to purchase an additional 900 megawatts. The company said that this option was exercised earlier this week, which could significantly expand the scope of Bloom Energy's involvement in the project and support the planned capacity of the Wyoming facility.

AEP also disclosed that it has signed a 20-year offtake arrangement with an unnamed high investment grade third-party customer for 100% of the facility's output, subject to certain conditions that the company expects could be satisfied by the second quarter of 2026. 

My video explains key fundamental pivots like their partnership with Brookfield, how BE could easily exceed the revenue-doubling of Generac ((GNRC Quick QuoteGNRC - Free Report) ), and why nuclear SMRs (small modular reactors) were still a very distant solution (dream?).

I especially enjoyed highlighting my view that Brookfield CEO Bruce Flatt is the "Dan Loeb" of infrastructure investing who you definitely want to have in your corner!

I also go over all the amazing growth metrics on the top and bottom lines which make Bloom Energy a $350 stock in the next 8-12 months if they keep executing the way they have.

Bloom 2026 Data Center Power Report: The New Realities Shaping AI Buildout

Management just updated their fundamental views of the market and here's their executive summary...

Early this year, our 2026 Data Center Power Report identified power availability as the defining constraint on data center growth. This mid-year update finds that power remains the dominant issue. However, other challenges are increasingly affecting the buildout of large-scale projects, hampering the speed of execution.

Our research among 156 data center decision-makers includes hyperscalers, colocation providers, neoclouds, data center developers, and chip developers, and is supplemented by public announcements and conversations with industry leaders.

The findings point to an industry that remains on track for significant growth, but one in which competitive advantage hinges on securing power, navigating permitting, earning community support, managing emissions, and deploying next-generation architectures. 

As AI infrastructure scales, leadership will come down to addressing all of these requirements at speed and scale.

1) A prolonged expansion of data center capacity is underway. US data center electricity demand is projected to more than double by 2030, and data center developers are planning an elevated pace of capacity additions through the end of the decade. The composition of AI workloads is changing faster than expected, with inference already accounting for over half of AI compute today. This shift reflects AI’s transition from model building to real-world applications, with inference workloads driving sustained demand for new data center capacity.

2) Power remains the biggest challenge to bringing new capacity online, but other barriers are gaining importance. While access to power is still the dominant issue for data center development, construction costs and community scrutiny have emerged as growing barriers. Developers identify higher local electricity prices, increased water consumption, and strain on grid reliability as the community concerns most likely to influence projects. Solutions that reduce local impacts are becoming increasingly critical to project success. 

3) Carbon capture is moving from concept to deployment as developers look to reconcile rapid power growth with emissions reduction goals. By 2030, nearly one-third of US data center sites using onsite power are expected to incorporate carbon capture, utilization, and storage (CCUS), reaching more than 40% by 2035. This planned adoption reflects growing pressure to expand power capacity while addressing emissions concerns.

4) The AC-to-DC transition is advancing faster than expected, creating a growing readiness gap. As higher rack densities drive new power delivery requirements, chip developers expect hybrid AC-DC architectures to be adopted in 2028, a full year ahead of data center developers’ plans. DC-native designs will follow quickly, accounting for 36% of new deployments in less
than four years. The architecture that developers choose today will determine whether they can support the next generation of AI chips.

Bottom line on Bloom: If you watch my video, you'll see where I have identified key "buy zones" for Bloom Energy (BE Quick QuoteBE - Free Report) in May and June and these will continue to be "higher-lows" of support as the stock ascends to and sustains above $300 into Q3 and Q4.