Bloom Energy recently reported strong second-quarter results at a time when demand for electricity from AI data centers is rising rapidly. Artificial intelligence is reshaping more than software: it is also changing how technology companies think about energy, infrastructure and the speed at which new computing capacity can come online.
Data centers are becoming larger and more power-intensive, while connecting new facilities to the electrical grid can involve long waits. That combination is increasing interest in onsite generation as businesses look for ways to secure dependable power without relying entirely on the pace of traditional grid expansion.
AI is contributing to a sharp increase in data-center power demand
The rapid expansion of artificial intelligence has created an energy challenge unlike anything the technology sector has faced before. Training large language models, operating cloud computing platforms and supporting AI applications all require substantial amounts of electricity.
At the same time, utility interconnection queues can delay new projects by months or even years. For data center operators, securing enough power has therefore become a major part of the growth equation rather than a background infrastructure issue.
Bloom Energy’s forecast suggests that by 2030, approximately 30% of data center sites could use onsite power as a primary energy source alongside the grid. The projection points to a broader shift in energy strategy: rather than depending exclusively on utility infrastructure, operators are increasingly considering generation located at or near the facility to gain more control over reliability and deployment timelines.
The Quarter That Confirmed a Market Shift
Bloom Energy's second-quarter results arrived against this backdrop of growing demand for onsite power. The company reported revenue of $1.07 billion, more than doubling the $401 million generated during the same quarter last year and exceeding analyst expectations. Adjusted earnings also exceeded expectations, while profitability, free cash flow and operating margin were all higher than a year earlier.
Bloom also raised its full-year revenue guidance to between $3.9 billion and $4.2 billion, indicating that management expects demand to remain strong beyond a single quarter.
During an earnings call, KR Sridhar, founder, chairman and chief executive officer of Bloom Energy, said the company’s backlog is expanding as customers rethink how they secure power for new infrastructure.
“The demand for Bloom Energy’s solutions keeps accelerating every quarter as customers who traditionally defaulted to combustion technologies are now proactively choosing Bloom as a superior power solution. Today, all the major US hyperscalers and over a dozen US neoclouds, AI labs, and colocation data center operators have validated and approved our power solutions for their AI factories. Bloom is now a standard for AI onsite power,” Sridhar stated.
Why Data Centers Need a New Energy Strategy
For large data centers, reliability is critical. Outages can create substantial operational and financial consequences, while companies are also under pressure to bring new facilities online quickly and manage the environmental impact of growing electricity consumption.
Bloom Energy addresses that need through onsite solid oxide fuel cells. Its Energy Servers convert fuels such as natural gas, biogas and hydrogen directly into electricity through an electrochemical process rather than conventional combustion. The company positions this approach as a way to reduce local air pollutants and particulate matter while lowering carbon emissions compared with some conventional forms of power generation.
The significance for data center operators is not simply the technology itself. It is the possibility of adding power close to where demand is created, reducing dependence on the timing of major grid upgrades and giving operators another option when utility capacity is constrained.
Reliability Is Becoming Just as Important as Sustainability
The AI economy depends on uninterrupted power. Bloom Energy says its onsite generation platform can provide availability of up to 99.999%. Because Energy Servers generate electricity continuously and can use the underground natural-gas pipeline network, the company argues that the systems can provide greater resilience during grid outages and other disruptions than approaches that rely entirely on utility supply or conventional backup generation.
The systems are also designed to scale from deployments of around 20 megawatts to 500 megawatts and beyond. Their modular structure allows additional capacity to be added as demand grows, potentially giving operators more flexibility than waiting for major utility infrastructure projects to be completed.
That combination of resilience, scalability and deployment speed helps explain why onsite generation is increasingly being considered as part of primary power infrastructure rather than solely as emergency backup.
Looking Beyond the Numbers
Bloom Energy's recent performance is notable because it coincides with a wider change in how data center operators think about electricity. AI infrastructure is expanding at a pace that can outstrip the speed at which new grid capacity becomes available, forcing companies to consider a broader range of power strategies.
Onsite generation is one of those strategies. It does not remove the need for utility infrastructure, nor does one strong quarter establish which technology will ultimately dominate the market. But it does show that access to reliable power is becoming a competitive issue for companies building AI infrastructure.
Bloom's second-quarter results therefore matter beyond the company itself. They offer a snapshot of a market in which electricity availability, deployment speed and resilience are becoming central considerations in the next phase of data center growth.