A new wave of artificial intelligence innovation is set to sharply increase global electricity demand, with data centre power use projected to rise by 160% by 2030.
Currently consuming 1–2% of global electricity, data centres could account for 3–4% by decade’s end as AI workloads surge.
The primary driver is the computational intensity of AI models — training and running them require far more power than traditional workloads. A single large-language-model query can use nearly 10 times more electricity than a standard search.
AI-focused data centres are forecast to grow their energy use by 30% annually, compared to 9% for conventional servers. This surge poses major challenges for power grids and environmental sustainability.
In the U.S., data centres could represent nearly half of future electricity demand growth if trends continue.
Utilities must expand capacity and integrate more renewable energy, while operators adopt efficient cooling and design practices.
While AI can help improve energy optimization and clean-tech innovation, experts caution that un-checked growth could offset recent gains in efficiency.
As global competition for computing power intensifies, balancing performance with sustainability will define the next phase of the AI revolution.
See What’s Next in Tech With the Fast Forward Newsletter
Tweets From @varindiamag
Nothing to see here - yet
When they Tweet, their Tweets will show up here.
