As the scale of artificial intelligence (AI) development continues to expand, electricity is emerging as the next critical bottleneck. This week, chip design company Broadcom announced the establishment of a financing platform together with investment giants Apollo Global Management and Blackstone, aiming to support over 20 gigawatts of AI computing capacity by 2028, with an initial funding scale of $35 billion. Even automakers want in on the action. General Motors stated this week that it is developing a sodium-ion battery intended for energy storage in data centers and power grids.
All of this demand is falling on power grid systems that take years to upgrade. Large-scale data center campuses may have to wait many years to obtain grid connections, while the equipment needed to build new power plants is also in short supply. This mismatch between supply and demand is precisely the opportunity for three companies: Bloom Energy (BE), GE Vernova (GEV), and Vistra (VST).
Bloom Energy produces solid oxide fuel cells—systems that generate electricity from natural gas on-site without combustion, allowing data centers to bypass waiting for grid connections. Bloom Energy’s first-quarter revenue grew 130% year-over-year to $751.1 million, with product revenue increasing by 208%. The company also recorded a net profit of $70.7 million, turning around from a loss in the previous year. Management raised its full-year performance guidance and currently expects 2026 revenue between $3.4 billion and $3.8 billion, representing approximately 80% growth at the midpoint, up from the previous guidance of about 60%. In April, Oracle announced that Bloom Energy’s fuel cells would provide all the power for its AI data center campus, Project Jupiter, in New Mexico, with a capacity of up to 2.45 gigawatts, replacing the originally planned gas turbines and diesel generators. Notably, management stated that more than half of Bloom Energy’s data center backlog comes from customers other than Oracle.
While Bloom Energy helps data centers bypass the grid, GE Vernova supplies equipment for the grid itself and for the gas turbines that utilities are queuing up to order. The power equipment manufacturer’s first-quarter organic orders surged 71% to $18.3 billion, pushing its total backlog to $163 billion. Its gas turbine backlog and slot reservation agreements grew from 83 gigawatts to 100 gigawatts in a single quarter, and management now expects at least 110 gigawatts by the end of 2026. GE Vernova’s electrification division, which produces grid equipment such as transformers and switchgear, signed $2.4 billion in equipment orders to support data centers in the first quarter, exceeding the total for all of 2025. Additionally, the company’s free cash flow more than quadrupled year-over-year to $4.8 billion, and management also raised its 2026 performance guidance.
Vistra is one of the largest competitive power producers in the United States, with generation assets encompassing natural gas and nuclear energy. The biggest spenders in AI are locking in their power generation capacity years in advance. Last year, the company signed a 20-year power purchase agreement with Amazon’s cloud division to secure up to 1,200 megawatts of nuclear power from its Comanche Peak nuclear plant in Texas. In addition, Vistra signed a 20-year agreement with Meta Platforms to supply 2,609 megawatts of nuclear power and generation capacity from its Eastern U.S. power plants. The company is also increasing its natural gas generation capacity, planning to acquire approximately 5,500 megawatts of generation assets from Cogentrix, with the deal expected to close in the second half of this year. Last month, Vistra reported first-quarter adjusted EBITDA from continuing operations of approximately $1.5 billion and reaffirmed its full-year performance expectation of $6.8 billion to $7.6 billion.
Of course, these stocks also face risks. For example, project timeline risks—Bloom Energy’s stock fell this week because a partner paused work on a data center site in Wyoming. Additionally, all of these companies operate in highly regulated markets. A larger risk may come from the demand side itself. All three stocks have risen based on the assumption of continued growth in AI capital expenditures, and recently, AI infrastructure stocks have already pulled back due to market concerns about the pace of spending. If AI construction slows significantly, backlogs could stop growing, and valuations could quickly compress.
AI Financial Service Technology Utilities
We connect Investors to Opportunities.
Financial Market Data copyright © 2020 QuoteMedia. Data delayed 15 minutes unless otherwise indicated (view delay times for all exchanges). RT=Real-Time, EOD=End of Day, PD=Previous Day. Market Data powered by QuoteMedia. Terms of Use.
© 2026 NAI Interactive Ltd. All Rights Reserved
Website Design Vancouver by Nirvana Canada
Leave a Reply