Archive — JP — japantimes
AI surge to double data center electricity demand by 2030, report says
A report from the IEA projects that electricity consumption by data centers will more than double by 2030 due to AI applications, raising energy security and CO2 emission concerns. Despite this, AI could improve energy efficiency. The U.S., Europe, and China currently account for most of the consumption, with major companies like Google and Microsoft investing in nuclear power for data centers.
Key points
- Data centers currently consume about 1.5% of global electricity (2024), increasing by 12% annually.
- AI applications are identified as the primary driver for projected doubling of electricity demand, reaching 945 TWh by 2030.
- The U.S., Europe, and China represent approximately 85% of total data center consumption.
- Major tech firms, including Google, Microsoft, and Amazon, are pivoting to nuclear energy sources to meet rising power needs.
- Data centers may consume about 3% of global energy by 2030, significantly raising carbon emissions linked to electricity use (projected increase from 180 million tonnes to 300 million tonnes by 2035).
Why it matters
- The rise of AI coincides with historical trends in data consumption and storage, reminiscent of the internet boom.
- Current trends suggest a shift towards renewable energy sources, resembling the early 21st-century transition from coal to cleaner energy.
- Concerns over energy security and carbon emissions align with global climate initiatives aimed at reducing overall emissions.
The IEA highlights a critical intersection of AI growth and data center energy consumption, indicating a pressing need for sustainable energy solutions that balance increasing demand with environmental concerns by 2030.