AI will spike data centre power demand 160% by 2030
A single ChatGPT query consumes nearly 10 times the electricity of a Google search
Increased use of AI is set to drive a dramatic surge in data centre power demand, with a projected increase of 160% by 2030, according to new research.
This surge is expected to have significant implications for power consumption and costs globally, particularly in the US and Europe, says Goldman Sachs Research.
Datacentres, which currently consume 1-2% of the world's electricity, are predicted to see this share rise to 3-4% by the end of the decade. In the US and Europe, this increased demand marks a return to electricity consumption growth not seen in a generation, with the potential for datacentre carbon dioxide emissions to more than double from 2022 levels by 2030.
The energy appetite of AI
A key driver behind this surge is the energy required for AI processing. A single ChatGPT query consumes nearly 10 times the electricity of a Google search (2.9Wh versus 0.3 Wh, according to the International Energy Agency), highlighting the broader impact of AI on energy consumption.
Goldman Sachs estimates that AI alone will contribute an additional 200 terawatt-hours per year to datacentre power consumption between 2023 and 2030.
Some AI innovations boost computing speed faster than they ramp up their electricity use, but the widening use of AI will still imply an increase in the technology's consumption of power, the research found.
Read more: Power demand from UK datacentres set to surge six-fold over the next decade
US and European power demand
In the US, datacentres are set to account for 8% of the nation's power consumption by 2030, up from 3% in 2022. This increase will necessitate around $50 billion in new generation capacity investment and an estimated 3.3 billion cubic feet per day of new natural gas demand. The last significant rise in US power demand occurred in the early 2000s, making this a notable shift.
Europe, meanwhile, faces a different set of challenges and opportunities. Having experienced a decline in electricity demand since 2008, Europe is now poised for a resurgence, with power demand potentially increasing by up to 50% from 2023 to 2033. This growth will be driven by datacentres and accelerated electrification. Meeting these needs will require over €1 trillion in investment to upgrade Europe's ageing power grid and expand renewable energy sources.
Future implications
The Goldman Sachs analysts emphasise that, while technology companies remain confident in reducing energy intensity, they are less certain about meeting absolute emissions targets due to rising demand. This situation highlights the importance of substantial investments in renewables and emerging nuclear technologies.
Despite the challenges, AI also holds the potential to drive innovation in areas such as healthcare, agriculture and energy efficiency, which could mitigate some of the environmental impacts, Goldman Sachs says.
Datacentres' massive demand growth is already causing shifts in the energy landscape. Today a utility company in Ohio asked the state's public utilities commission to approve proposals creating a new customer class and tariffs for datacentres, protecting other users from rate rises.