Skip to main content

Onsite Generation Expected to Fully Power 27% of Data Center Facilities by 2030

Data centers are adopting onsite power as a primary energy source, according to the 2025 Data Center Power Report from Bloom Energy.

Data centers are likely to continue to struggle with the timely availability of electricity, according to the report, which carries major implications for the future of the AI industry.

The report's mid-year update shows that securing electricity for data centers is likely to take much longer than anticipated, and that power availability is now the leading factor in site selection. The report offers a timely lens into what matters most to the leaders shaping the future of the AI industry in America, including:

Data center developers are underestimating time to power

Utility providers report significantly longer timelines to deliver power in key US markets, up to 2 years longer than what hyperscalers and colocation providers expect.

Power access is a leading factor in data center site selection

84% of respondents ranked availability of power among their top three considerations.

Onsite power is increasingly critical

In 2030, 38% of facilities are expected to use some onsite generation for primary power, up from 13% a year ago. Notably, 27% of facilities expect to be fully powered by onsite generation by 2030, a 27x increase from just 1% last year.

AI is driving larger, more power-intensive data centers

The median data center size is expected to grow by nearly 115%, from approximately 175 MW today to about 375 MW over the next 10 years.

Reducing carbon emissions is a lower but lasting priority

95% of those surveyed affirmed that sustainability and carbon reduction targets are still in place, even if the path to achieving those goals may not be linear.

"Decisions around where data centers get built have shifted dramatically over the last six months, with access to power now playing the most significant role in location scouting," said Aman Joshi, Bloom Energy's Chief Commercial Officer. "The grid can't keep pace with AI demands, so the industry is taking control with onsite power generation. When you control your power, you control your timeline, and immediate access to energy is what separates viable projects from stalled ones."

According to the survey, operators are looking beyond legacy power generation to solutions that offer fast deployment timelines, low emissions, and the ability to handle intense and fluctuating AI workloads, all while meeting the industry's uncompromising reliability standards and cost requirements.

Methodology: The latest report is based on data collected from April 2024 to April 2025, which surveyed approximately 100 decision-makers across the entire data center power ecosystem, reflecting perspectives from hyperscalers, colocation developers, utilities, and GPU service providers. 

Hot Topics

The Latest

Payment system failures are putting $44.4 billion in US retail and hospitality sales at risk each year, underscoring how quickly disruption can derail day-to-day trading, according to research conducted by Dynatrace ... The findings show that payment failures are no longer isolated incidents, but part of a recurring operational challenge that disrupts service, damages customer trust, and negatively impacts revenue ...

For years, the success of DevOps has been measured by how much manual work teams can automate ... I believe that in 2026, the definition of DevOps success is going to expand significantly. The era of automation is giving way to the era of intelligent delivery, in which AI doesn't just accelerate pipelines, it understands them. With open observability connecting signals end-to-end across those tools, teams can build closed-loop systems that don't just move faster, but learn, adapt, and take action autonomously with confidence ...

The conversation around AI in the enterprise has officially shifted from "if" to "how fast." But according to the State of Network Operations 2026 report from Broadcom, most organizations are unknowingly building their AI strategies on sand. The data is clear: CIOs and network teams are putting the cart before the horse. AI cannot improve what the network cannot see, predict issues without historical context, automate processes that aren't standardized, or recommend fixes when the underlying telemetry is incomplete. If AI is the brain, then network observability is the nervous system that makes intelligent action possible ...

SolarWinds data shows that one in three DBAs are contemplating leaving their positions — a striking indicator of workforce pressure in this role. This is likely due to the technical and interpersonal frustrations plaguing today's DBAs. Hybrid IT environments provide widespread organizational benefits but also present growing complexity. Simultaneously, AI presents a paradox of benefits and pain points ...

Over the last year, we've seen enterprises stop treating AI as “special projects.” It is no longer confined to pilots or side experiments. AI is now embedded in production, shaping decisions, powering new business models, and changing how employees and customers experience work every day. So, the debate of "should we adopt AI" is settled. The real question is how quickly and how deeply it can be applied ...

In MEAN TIME TO INSIGHT Episode 20, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA presents his 2026 NetOps predictions ... 

Today, technology buyers don't suffer from a lack of information but an abundance of it. They need a trusted partner to help them navigate this information environment ...

My latest title for O'Reilly, The Rise of Logical Data Management, was an eye-opener for me. I'd never heard of "logical data management," even though it's been around for several years, but it makes some extraordinary promises, like the ability to manage data without having to first move it into a consolidated repository, which changes everything. Now, with the demands of AI and other modern use cases, logical data management is on the rise, so it's "new" to many. Here, I'd like to introduce you to it and explain how it works ...

APMdigest's Predictions Series continues with 2026 Data Center Predictions — industry experts offer predictions on how data centers will evolve and impact business in 2026 ...

APMdigest's Predictions Series continues with 2026 DataOps Predictions — industry experts offer predictions on how DataOps and related technologies will evolve and impact business in 2026. Part 2 covers data and data platforms ...

Onsite Generation Expected to Fully Power 27% of Data Center Facilities by 2030

Data centers are adopting onsite power as a primary energy source, according to the 2025 Data Center Power Report from Bloom Energy.

Data centers are likely to continue to struggle with the timely availability of electricity, according to the report, which carries major implications for the future of the AI industry.

The report's mid-year update shows that securing electricity for data centers is likely to take much longer than anticipated, and that power availability is now the leading factor in site selection. The report offers a timely lens into what matters most to the leaders shaping the future of the AI industry in America, including:

Data center developers are underestimating time to power

Utility providers report significantly longer timelines to deliver power in key US markets, up to 2 years longer than what hyperscalers and colocation providers expect.

Power access is a leading factor in data center site selection

84% of respondents ranked availability of power among their top three considerations.

Onsite power is increasingly critical

In 2030, 38% of facilities are expected to use some onsite generation for primary power, up from 13% a year ago. Notably, 27% of facilities expect to be fully powered by onsite generation by 2030, a 27x increase from just 1% last year.

AI is driving larger, more power-intensive data centers

The median data center size is expected to grow by nearly 115%, from approximately 175 MW today to about 375 MW over the next 10 years.

Reducing carbon emissions is a lower but lasting priority

95% of those surveyed affirmed that sustainability and carbon reduction targets are still in place, even if the path to achieving those goals may not be linear.

"Decisions around where data centers get built have shifted dramatically over the last six months, with access to power now playing the most significant role in location scouting," said Aman Joshi, Bloom Energy's Chief Commercial Officer. "The grid can't keep pace with AI demands, so the industry is taking control with onsite power generation. When you control your power, you control your timeline, and immediate access to energy is what separates viable projects from stalled ones."

According to the survey, operators are looking beyond legacy power generation to solutions that offer fast deployment timelines, low emissions, and the ability to handle intense and fluctuating AI workloads, all while meeting the industry's uncompromising reliability standards and cost requirements.

Methodology: The latest report is based on data collected from April 2024 to April 2025, which surveyed approximately 100 decision-makers across the entire data center power ecosystem, reflecting perspectives from hyperscalers, colocation developers, utilities, and GPU service providers. 

Hot Topics

The Latest

Payment system failures are putting $44.4 billion in US retail and hospitality sales at risk each year, underscoring how quickly disruption can derail day-to-day trading, according to research conducted by Dynatrace ... The findings show that payment failures are no longer isolated incidents, but part of a recurring operational challenge that disrupts service, damages customer trust, and negatively impacts revenue ...

For years, the success of DevOps has been measured by how much manual work teams can automate ... I believe that in 2026, the definition of DevOps success is going to expand significantly. The era of automation is giving way to the era of intelligent delivery, in which AI doesn't just accelerate pipelines, it understands them. With open observability connecting signals end-to-end across those tools, teams can build closed-loop systems that don't just move faster, but learn, adapt, and take action autonomously with confidence ...

The conversation around AI in the enterprise has officially shifted from "if" to "how fast." But according to the State of Network Operations 2026 report from Broadcom, most organizations are unknowingly building their AI strategies on sand. The data is clear: CIOs and network teams are putting the cart before the horse. AI cannot improve what the network cannot see, predict issues without historical context, automate processes that aren't standardized, or recommend fixes when the underlying telemetry is incomplete. If AI is the brain, then network observability is the nervous system that makes intelligent action possible ...

SolarWinds data shows that one in three DBAs are contemplating leaving their positions — a striking indicator of workforce pressure in this role. This is likely due to the technical and interpersonal frustrations plaguing today's DBAs. Hybrid IT environments provide widespread organizational benefits but also present growing complexity. Simultaneously, AI presents a paradox of benefits and pain points ...

Over the last year, we've seen enterprises stop treating AI as “special projects.” It is no longer confined to pilots or side experiments. AI is now embedded in production, shaping decisions, powering new business models, and changing how employees and customers experience work every day. So, the debate of "should we adopt AI" is settled. The real question is how quickly and how deeply it can be applied ...

In MEAN TIME TO INSIGHT Episode 20, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA presents his 2026 NetOps predictions ... 

Today, technology buyers don't suffer from a lack of information but an abundance of it. They need a trusted partner to help them navigate this information environment ...

My latest title for O'Reilly, The Rise of Logical Data Management, was an eye-opener for me. I'd never heard of "logical data management," even though it's been around for several years, but it makes some extraordinary promises, like the ability to manage data without having to first move it into a consolidated repository, which changes everything. Now, with the demands of AI and other modern use cases, logical data management is on the rise, so it's "new" to many. Here, I'd like to introduce you to it and explain how it works ...

APMdigest's Predictions Series continues with 2026 Data Center Predictions — industry experts offer predictions on how data centers will evolve and impact business in 2026 ...

APMdigest's Predictions Series continues with 2026 DataOps Predictions — industry experts offer predictions on how DataOps and related technologies will evolve and impact business in 2026. Part 2 covers data and data platforms ...