RESEARCH
AI-Driven Data Centers on Track to Nearly Double Global Electricity Consumption by 2030
Multiple independent energy agencies and research institutions project global data center electricity consumption will nearly double — from roughly 415 TWh in 2024 to 945 TWh by 2030 — driven primarily by accelerating AI workload growth. In the US, data centers could account for 8–12% of total electricity demand by 2030, up from approximately 4% today. Grid operators including NERC are already revising long-term reliability forecasts upward to account for these new industrial loads. The convergence of government agencies, financial institutions, and grid operators signals this is no longer speculative: it is an infrastructure planning imperative.
Findings
- IEA projects global data center electricity will grow from 415 TWh in 2024 to 945 TWh by 2030 — a 128% increase — with AI-focused accelerated servers growing 30% annually
- Berkeley Lab and DOE estimate US data center consumption could reach 325–580 TWh by 2028, representing 6.7–12% of total US electricity, up from 4.4% in 2023
- Goldman Sachs projects a 165% increase in global data center power demand by 2030, with AI workloads comprising 39% of total data center demand
- NERC 2025 Long-Term Reliability Assessment forecasts 224 GW of new US summer peak demand over 10 years, with data centers and AI as the dominant growth driver
- McKinsey estimates US data center power demand will more than triple — from 25 GW in 2024 to 70–80 GW by 2030 — requiring significant new grid infrastructure investment
Delta Engine result
↔ Divergence Detected — Δ 0.0851 (threshold 0.05)
- IEA T5 95
- Lawrence Berkeley National Laboratory T5 92
- Goldman Sachs Research T0 98
- McKinsey & Company T0 96
- NERC T0 90
Evidence quality
avg 53 · min 25 · max 95 · spread 70