Green cloud computing: data center power use is set to double by 2030
Data centers used about 1.5% of the world's electricity in 2024. The International Energy Agency expects that to more than double by 2030, with AI the biggest driver.
The International Energy Agency's 2025 report Energy and AI put numbers on a question that used to be argued in generalities. Data centers used about 415 terawatt-hours of electricity in 2024, roughly 1.5% of the world's total. The IEA projects that figure will reach around 945 terawatt-hours by 2030, more than double, and names AI as the most important driver of the growth.
The United States accounted for 45% of global data center electricity use in 2024, according to the report.
Why IT buyers should care
For most companies, data center energy is something their cloud provider buys, not them. It still reaches IT in three ways.
First, cost. Energy is a large part of what it costs to run a data center, and rising demand puts pressure on prices for compute, especially GPU capacity.
Second, reporting. Companies with emissions reporting duties often have to account for the emissions of the cloud services they use, as part of their supply chain.
Third, capacity. Power availability is now a constraint on where new data centers can be built, which affects the regions and capacity available to customers.
How to reduce the footprint of your cloud use
Most of the practical steps also cut costs:
| Step | Effect |
|---|---|
| Switch off idle and oversized resources | The cheapest kilowatt-hour is the one not used |
| Use autoscaling | Capacity follows demand instead of sitting at peak |
| Pick the right region | Providers publish carbon data by region; some regions run on cleaner grids |
| Prefer managed services | Shared platforms usually run at higher utilization than single-tenant servers |
| Right-size AI workloads | A smaller model that meets the need uses far less compute than the largest one |
| Delete data you don't need | Storage and its replicas consume power indefinitely |
The large providers each offer a carbon footprint tool in their consoles, such as AWS's Customer Carbon Footprint Tool, Microsoft's Emissions Impact Dashboard and Google Cloud's Carbon Footprint. They are a reasonable starting point for reporting, though methods differ between providers.
What IT teams should do
- Turn on your cloud provider's carbon footprint reporting.
- Check whether your company's emissions reporting covers cloud services, and who owns that data.
- Add idle-resource clean-up to your monthly cost review, so savings and emissions cuts are tracked together.
- When choosing AI models, test whether a smaller model meets the requirement.