 |
| Cover Image |
Key Takeaways
Global data center electricity consumption reached approximately 415 TWh in 2024, accounting for about 1.5% of global electricity demand.
According to the International Energy Agency (IEA), electricity consumption from data centers is projected to more than double to around 945 TWh by 2030.
AI-accelerated servers are expected to grow at 30% annually, becoming the primary driver of electricity demand growth.
Goldman Sachs Research forecasts that U.S. data center power demand will more than double from 31 GW in 2025 to 66 GW in 2027.
EPRI estimates data centers could consume 9% to 17% of total U.S. electricity generation by 2030.
Regional power grids, particularly in Virginia, the Mid-Atlantic, and parts of Asia and Europe, face increasing strain due to AI infrastructure expansion.
The future of AI power demand will depend heavily on efficiency improvements, renewable energy deployment, and the pace of AI adoption.
AI Data Center Electricity Demand Forecast Dashboard
415 TWh
Global Data Center Electricity Use (2024)
945 TWh
IEA Forecast for 2030
30%
Annual AI Server Growth
66 GW
US Data Center Power Demand (2027)
The Starting Point for Future Forecasts
The International Energy Agency's (IEA) 2026 report Energy and AI provides one of the most comprehensive assessments of global data center electricity use. According to the agency, data centers consumed approximately 415 terawatt-hours (TWh) of electricity in 2024, representing around 1.5% of total global electricity demand.
While this figure may appear modest compared to industrial, transportation, or residential consumption, the growth trajectory is what concerns policymakers, utilities, and investors. The IEA notes that electricity demand from data centers has increased at an annual rate of approximately 12% over the last five years, making it one of the fastest-growing electricity-consuming sectors worldwide.
The underlying composition of data center electricity consumption is also changing. Traditional enterprise servers are gradually being replaced by accelerated computing infrastructure powered by GPUs and AI accelerators. These systems consume substantially more power but are essential for training and operating large language models, generative AI platforms, and advanced machine learning applications.
As a result, future electricity demand forecasts increasingly depend not on the number of data centers being built, but on the intensity of AI workloads running inside them.
IEA Forecast: Global Data Center Electricity Demand Could More Than Double by 2030
Among all available projections, the International Energy Agency's 2026 forecast remains one of the most widely referenced by governments, utilities, and energy planners.
In the agency's Base Case scenario, global electricity consumption from data centers rises from 415 TWh in 2024 to approximately 945 TWh by 2030, representing annual growth of roughly 15% per year.
The significance of this projection extends beyond the absolute numbers. According to the IEA, data center demand is expected to grow more than four times faster than electricity demand from all other sectors combined during the same period.
The report identifies AI as the primary driver of this acceleration. Electricity consumption from accelerated AI servers is projected to increase by approximately 30% annually, accounting for nearly half of total data center electricity demand growth through 2030.
Dr. Fatih Birol, Executive Director of the International Energy Agency, has repeatedly emphasized that artificial intelligence is rapidly becoming a major force shaping future energy systems. The IEA's findings suggest that AI infrastructure will become an increasingly important factor in national electricity planning, grid modernization strategies, and energy investment decisions.
By 2030, data centers are expected to account for nearly 3% of total global electricity consumption, compared to 1.5% in 2024.
IEA Global Electricity Demand Forecast
| Year | Electricity Demand (TWh) | Share of Global Electricity |
|---|---|---|
| 2024 | 415 | 1.5% |
| 2030 | 945 | ~3% |
| 2035 (Base) | 1,200+ | ~3.5% |
Goldman Sachs Forecast Signals an AI Infrastructure Boom
While the IEA focuses on global energy systems, Goldman Sachs Research has concentrated heavily on the impact of AI infrastructure expansion in the United States.
In the May 2026 research note US Data Center Power Demand Projected to Double by 2027, commodities analysts Hongcen Wei, Daan Struyven, and Samantha Dart forecast a dramatic acceleration in electricity demand.
According to their analysis, U.S. data center power demand is expected to increase from approximately 31 gigawatts (GW) in 2025 to 41 GW in 2026 and 66 GW in 2027.
The forecast is based on detailed facility-level tracking, permitting data, construction activity, and satellite imagery compiled by infrastructure intelligence provider Aterio.
What makes this forecast particularly significant is the speed of the expansion. The researchers estimate that U.S. data center capacity could exceed 95 GW by the end of 2027, more than doubling within just two years.
The report also warns that actual growth could vary considerably depending on project execution. Historically, only about 72% of planned facilities have entered service on schedule, with labor shortages, equipment constraints, permitting delays, and utility interconnection challenges frequently disrupting deployment timelines.
Nevertheless, even after applying conservative adjustments, Goldman Sachs expects AI-driven infrastructure growth to remain one of the largest sources of incremental electricity demand in North America.
Goldman Sachs U.S. Forecast
Which Regions Will See the Largest Growth?
The International Energy Agency expects the United States, China, and Europe to remain the largest markets. These states alone account for nearly 80% of projected global data center electricity demand growth through 2030.
United States
The IEA projects U.S. data center electricity consumption to increase by approximately 240 TWh, representing growth of around 130% compared with 2024 levels.
The agency also estimates that U.S. per-capita data center electricity consumption could exceed 1,200 kWh annually by 2030, the highest level globally.
China
China is expected to experience even faster growth.
According to the IEA, electricity consumption from data centers could increase by approximately 175 TWh, representing growth of around 170% through 2030.
Europe
European data center electricity demand is projected to rise by more than 45 TWh, equivalent to roughly 70% growth by the end of the decade.
Southeast Asia
The IEA identifies Southeast Asia as one of the fastest-growing emerging markets, with electricity demand from data centers expected to more than double by 2030 due to rapid digitalization and regional cloud infrastructure expansion.
➡️ Also Read: AI Data Centers and the Global Electricity Surge: Why Power Is Becoming the New Bottleneck of Digital Infrastructure
| Region | Increase by 2030 | Growth Rate | Forecast Outlook |
|---|---|---|---|
| United States | +240 TWh | +130% | Very High |
| China | +175 TWh | +170% | Very High |
| Europe | +45 TWh | +70% | High |
| Southeast Asia | More Than Double | 100%+ | Fastest Emerging Market |
What do EPRI, McKinsey, Gartner, and BloombergNEF Offer About Forecast?
Several independent organizations have published forecasts that exceed the IEA's baseline outlook.
The Electric Power Research Institute (EPRI) estimated in February 2026 that data centers could account for between 9% and 17% of total U.S. electricity generation by 2030, substantially higher than earlier projections.
Particularly striking is EPRI's assessment of Virginia, the world's largest data center market. The organization projects that data centers could eventually consume between 39% and 57% of the state's electricity supply, compared with approximately 25% today.
McKinsey & Company similarly projects U.S. data center electricity consumption reaching approximately 606 TWh by 2030, representing nearly 11.7% of total U.S. power demand.
Meanwhile, Gartner's June 2026 market forecast estimates that global data center power demand will increase by 27% in a single year, rising from 104 GW in 2025 to approximately 132 GW in 2026.
BloombergNEF extends the timeline even further. Some long-term scenarios published by the research firm suggest global data center electricity demand could approach 1,200 TWh by 2035 and potentially 3,700 TWh by 2050 if AI adoption continues to accelerate.
Collectively, these forecasts illustrate how rapidly expectations have shifted. Just a few years ago, most energy models assumed efficiency improvements would largely offset growing computing demand. The emergence of generative AI has fundamentally changed that assumption.
| Organization | Forecast | Target Year |
|---|---|---|
| IEA | 945 TWh | 2030 |
| Goldman Sachs | 66 GW (US) | 2027 |
| EPRI | 9–17% of US electricity | 2030 |
| McKinsey | 606 TWh (US) | 2030 |
| BloombergNEF | 1,200 TWh | 2035 |
| BloombergNEF | 3,700 TWh | 2050 |
AI Servers Are Becoming the Dominant Driver of Electricity Demand
A central theme across every major forecast is the growing influence of accelerated computing infrastructure.
The IEA estimates that accelerated servers will account for nearly half of all future growth in data center electricity demand. Conventional servers, by contrast, are expected to contribute only around one-fifth of overall growth.
This distinction is important because AI servers consume significantly more electricity than traditional computing equipment.
Goldman Sachs Research notes that AI inference and model training workloads require specialized processors operating at extremely high utilization levels. As generative AI becomes embedded in search engines, productivity software, customer service systems, healthcare platforms, and industrial automation, demand for these accelerated systems is expected to expand dramatically.
By 2027, Gartner forecasts that AI-optimized servers will consume more electricity globally than conventional servers, marking a historic shift in the structure of data center energy demand.
➡️ Read Also: AI Data Center Power Crisis 2026: How It is Impacting U.S. Real Estate
Grid Capacity, Power Availability, and Infrastructure Constraints
The challenge facing utilities is not merely producing more electricity but delivering it where AI infrastructure is being built.
According to Goldman Sachs Research, reliability risks are particularly elevated across the Mid-Atlantic, Mid-Continent, and Northwest regions of the United States, where data center construction is occurring faster than generation capacity additions.
The researchers warn that power availability is increasingly becoming the primary factor influencing data center location decisions.
Similarly, the IEA emphasizes that data centers create unique planning challenges because electricity demand is highly concentrated geographically. A single hyperscale AI facility can consume as much electricity as a medium-sized city, requiring substantial investments in substations, transmission networks, backup systems, and generation resources.
These constraints are already influencing project timelines, utility planning processes, and corporate site-selection strategies worldwide.
Electricity Grid Stress Forecast
| Region | Risk Level |
|---|---|
| Virginia | Very High |
| Mid-Atlantic US | Very High |
| Mid-Continent US | High |
| Northwest US | High |
| Texas | Moderate |
| Georgia | Moderate |
How High Could Electricity Demand Rise?
Forecasting AI-driven electricity demand remains difficult because the technology itself is evolving rapidly.
To address this uncertainty, the IEA developed multiple scenarios.
Under the Base Case, demand reaches approximately 945 TWh by 2030.
The Lift-Off Scenario assumes faster AI adoption, improved supply chains, and fewer infrastructure bottlenecks. Under these conditions, global data center electricity demand could exceed 1,700 TWh by 2035, representing approximately 4.4% of global electricity consumption.
The High Efficiency Scenario assumes substantial improvements in chip performance, software optimization, cooling technologies, and facility design. In this case, electricity demand reaches approximately 970 TWh by 2035, despite strong AI adoption.
The Headwinds Scenario assumes slower AI deployment, tighter supply chains, permitting delays, and economic constraints. Under this outlook, electricity demand plateaus at roughly 700 TWh, limiting data centers to less than 2% of global electricity demand.
The wide gap between these scenarios demonstrates that technological efficiency gains and infrastructure investment decisions will play a crucial role in determining the ultimate energy footprint of artificial intelligence.
Why Forecasts Vary So Much
One reason forecasts differ significantly is uncertainty surrounding AI adoption.
The IEA uses several scenarios to account for this uncertainty.
Base Case
Approximately 945 TWh by 2030, reaching nearly 3% of global electricity demand.
Lift-Off Case
Assumes stronger AI adoption and fewer infrastructure bottlenecks.
Under this scenario, electricity demand could exceed 1,700 TWh by 2035, accounting for approximately 4.4% of global electricity demand.
High Efficiency Case
Assumes substantial advances in software optimization, chip efficiency, and cooling technologies.
Demand reaches around 970 TWh by 2035, significantly lower than the Lift-Off scenario.
Headwinds Case
Assumes slower AI adoption, permitting delays, supply-chain challenges, and infrastructure constraints.
Under this scenario, demand stabilizes at approximately 700 TWh, keeping data centers below 2% of global electricity demand.
These scenarios illustrate how technology improvements and policy decisions could dramatically alter future outcomes.
What This Means for the Energy Industry?
The AI boom is creating opportunities across the energy sector.
Utilities are investing in:
- Grid expansion
- Transmission infrastructure
- Renewable energy projects
- Battery storage
- Natural gas generation
- Small modular nuclear reactors (SMRs)
Meanwhile, technology companies are signing record-breaking power purchase agreements for renewable energy.
According to both the IEA and Goldman Sachs Research, AI data centers are becoming one of the most important sources of incremental electricity demand worldwide.
This trend is likely to influence energy markets, infrastructure planning, and investment decisions throughout the next decade.
Where Investment Is Flowing
The Forecast Points to Sustained Growth, Not a Temporary Surge
Virtually every major organization studying the intersection of artificial intelligence and energy—including the International Energy Agency, Goldman Sachs Research, EPRI, McKinsey, Gartner, BloombergNEF, and Lawrence Berkeley National Laboratory—reaches the same fundamental conclusion: AI is reshaping global electricity demand forecasts.
Although estimates vary, the direction is unmistakable. Data center electricity consumption is expected to grow significantly faster than overall electricity demand, with AI-accelerated computing becoming the dominant source of new load growth.
The debate is no longer whether AI will increase electricity demand. The real question is how quickly utilities, grid operators, regulators, and energy developers can build the infrastructure needed to support this unprecedented expansion. For investors, policymakers, and energy planners, AI data center electricity demand is no longer a niche technology issue—it has become a central component of future energy strategy.
Core Insights Review contributors publish research-based analysis and editorial insights on commercial real estate, PropTech, smart infrastructure, sustainable construction, industrial real estate, and emerging technologies shaping the future of the built environment.
Check for more information: Core Insights Review
Follow us at: LinkedIn, Facebook, and X
Post a Comment
0Comments