AI data centers now consume roughly 670,000 barrels of oil equivalent per day globally, and the energy demand is expected to more than double by 2030—triggering a competition for power that mirrors the fight over oil.

Yahoo Finance AI9h ago3 min read

Summaries like this, in your inbox every morning.

3 Key Points

What happened
Data centers pulled about 415 terawatt-hours off the world's grids in 2024, equivalent to roughly 670,000 barrels of oil equivalent a day. The International Energy Agency projects that by 2030, data-center electricity demand could surge toward 945 terawatt-hours, or about 1.5 million barrels of oil equivalent a day. Meanwhile, companies like Bitzero Holdings are using Bitcoin mining cash flow to build low-carbon power facilities for AI data centers, with a deal announced on May 5th to supply power from Norway facilities to OneQode Networks.
Why it matters
The energy appetite of AI infrastructure is now outpacing Bitcoin mining and competing with traditional energy-intensive industries. JLL estimates that meeting global data-center capacity needs by 2030 will require almost 100 gigawatts of new supply and as much as $3 trillion（約480兆円） in combined infrastructure and GPU spending. This scale of demand means AI companies are now scrambling for the same resource—secure access to energy—that has historically driven geopolitical competition.
What to watch
The daily output benchmark is striking: by 2030, AI data centers are expected to burn the daily output of a mid-sized oil producer just to train models and answer questions. JLL projects global data-center capacity will nearly double by 2030, making power capacity a critical bottleneck for the AI industry.

No discussion yet for this article

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.

Free · takes 30 seconds · unsubscribe anytime

5 minutes a day. The AI essentials.

200+ sources · Email / LINE / Slack