Last updated: June 2026. Reviewed by Report AI editorial. Every figure on this page is linked to its primary source.
AI INFRASTRUCTURE & COMPUTE 2026 — KEY DATA POINTS
415 TWh
Data-center electricity 2024
~1.5% of global power
IEA
~945 TWh
Projected by 2030
roughly doubling (~3% global)
IEA base case
+30%/yr
AI-server electricity growth
vs. 9%/yr conventional
IEA base case
Global data-center electricity (TWh), 2024–2035 (IEA, Energy and AI) · See all comparison charts →
Data centers consumed roughly 415 terawatt-hours of electricity in 2024 — about 1.5% of the global total — and demand is projected to roughly double to around 945 TWh by 2030 (IEA). Electricity for AI-accelerated servers is growing about 30% a year, far outpacing conventional servers. This page collects the most important primary-sourced AI infrastructure and compute statistics for 2026.
Executive summary
- Fast Fact: Data-center electricity use (~415 TWh in 2024, ~1.5% of global) has grown ~12% per year over the last five years and is set to roughly double by 2030.
- AI is the driver: Electricity for AI-accelerated servers is projected to grow ~30% annually, versus ~9% for conventional servers.
- Grid pressure: Gartner expects data-center electricity demand to grow 16% in 2025 and double by 2030; US data-center demand is projected to rise ~130% by 2030.
- Capital intensity: IDC forecasts AI infrastructure spending of $487B in 2026, exceeding $1 trillion by 2029.
Data-center electricity demand
The IEA’s Energy and AI analysis puts global data-center electricity at roughly 415 TWh in 2024 and projects it will reach around 945 TWh by 2030 in its base case — just under 3% of global electricity — climbing toward 1,200 TWh by 2035.
| Metric | Figure | Source |
|---|---|---|
| Data-center electricity (2024) | ~415 TWh (~1.5% global) | IEA |
| Projected (2030, base case) | ~945 TWh (~3% global) | IEA |
| Projected (2035) | ~1,200 TWh | IEA |
| AI-accelerated server growth | ~30% / year | IEA |
| Conventional server growth | ~9% / year | IEA |
| 5-year historical growth | ~12% / year | IEA |
Deep dive: the US grid strain
The pressure is concentrated where data centers cluster. The IEA estimates US data-center energy demand will rise about 130% by 2030; Lawrence Berkeley National Laboratory projects US data-center electricity climbing from 176 TWh in 2023 to a range of 325–580 TWh by 2028. Gartner separately forecasts global data-center electricity demand growing 16% in 2025 and doubling by 2030 (Gartner).
Infrastructure investment
The buildout is enormously capital-intensive. IDC forecasts AI infrastructure spending of about $487 billion in 2026 (~53% YoY), exceeding $1 trillion by 2029. For the full capital picture — venture funding, valuations, and hyperscaler capex — see our AI Investment & Funding Statistics 2026.
Frequently asked questions
How much electricity do data centers use?
About 415 TWh in 2024 — roughly 1.5% of global electricity — per the IEA, growing around 12% per year over the past five years.
How much will AI data-center energy grow?
The IEA projects global data-center electricity will roughly double to ~945 TWh by 2030 (just under 3% of global power), with AI-accelerated servers growing ~30% per year.
How much is being spent on AI infrastructure?
IDC forecasts AI infrastructure spending of roughly $487 billion in 2026 (~53% YoY growth), rising above $1 trillion by 2029.
Data sources & methodology
- International Energy Agency (IEA) — Energy and AI: Energy demand from AI, 2025.
Verified data points: ~415 TWh data-center electricity in 2024 (~1.5% global); ~945 TWh by 2030 (~3%); ~1,200 TWh by 2035; AI-server growth ~30%/yr vs ~9%/yr; ~12%/yr 5-year growth; US demand +130% by 2030.
Source: iea.org/reports/energy-and-ai - Gartner — data-center electricity forecast, November 2025.
Verified data points: Data-center electricity demand to grow 16% in 2025 and double by 2030.
Source: gartner.com - IDC — AI Infrastructure Spending tracker, 2026.
Verified data points: $487B AI infrastructure spend forecast for 2026 (~53% YoY); >$1T by 2029.
Source: idc.com
Related pages: AI Investment & Funding Statistics 2026 · What is AI Inference? · Methodology
Related Reports & Resources
Other reports in this cluster (Technology & Models): Generative AI Statistics 2026 — the workloads driving this energy demand · AI Model Benchmarks 2026 — the frontier-model race that’s setting compute requirements.
Compare year-over-year: AI infrastructure energy & spend chart · State of AI by year.
Background reading: AI Investment & Funding 2026 · AI Market Forecast 2026–2030 · Enterprise AI Statistics 2026.
Key concepts: AI Inference · Hyperscaler · GPU.
Browse the category: All Statistics & Reports → Technology & Models cluster.