AI Data Center Electricity Need To Double By 2030
The IEA has issued a new report titled “Energy and AI.” The authors claim it is the only global analysis examining every aspect of the junctions between energy and AI. This includes a view of how “sustainably meeting AI’s energy demand to how AI itself could transform the production, consumption, and transport of energy worldwide.”
The analysis looks at the regular list of possible energy-providing technologies ranging from wind to solar and nuclear to hydroelectric. Today advances in green energy are insufficient to offset the need for coal. The study also reviews that many energy grids are old and inadequate for rising demand.
A City Of 100,000 People
An AI data center can need an electricity supply comparable to that of a city of 100,000 people. In the future, some data centers will require as much as 20 times that level.
Last year, $500 billion was invested in data centers. About half of the electricity needed for these centers was in the US.
One of the study’s most important conclusions was that “Data centre electricity consumption is set to more than double to around 945 TWh by 2030.” That is roughly the amount used by the entire nation of Japan annually.
Residential Needs
The electricity competition is mostly for residential and legacy business needs. This will grow as electricity is needed to power Bitcoin mining and air conditioning.
Finally, the report does not have a good roadmap for how the electricity supply will grow enough in the next five years to meet anticipated demand.
More from ClimateCrisis 247
- Gas Prices Ready For Sharp Drop
- 3,521% Tariffs On Solar Energy
- Recession Could Kill Climate Rescue Advances
- No Energy Means No AI
