AI-focused data centres can require much more power than a traditional facility, but “AI” alone does not determine a site’s energy use, water demand or community impact. Capacity, workload, utilization, cooling design, climate and electricity supply all matter. The International Energy Agency (IEA) gives an illustrative scale contrast: traditional data centres commonly use 10–25 MW, while hyperscale AI centres can exceed 100 MW. Those are facility categories, not averages for every site.
How much more electricity does an AI data centre use?
There is no reliable single multiplier for “AI versus traditional.” An individual facility’s electricity demand depends on how much equipment it has, what it runs, how intensively it is used, and how much power its cooling and other support systems require.
Facility scale: a useful illustration, not a rule
| Facility category | Illustrative scale cited by the IEA | How to interpret it |
|---|---|---|
| Traditional data centre | Commonly 10–25 MW | An illustrative facility category, not a measured average across all traditional centres. |
| Hyperscale AI data centre | Can exceed 100 MW | A hyperscale category that may host AI workloads; it does not mean every AI facility is larger than every conventional one. |
These figures describe power scale, not annual electricity consumption. To compare two sites fairly, use their expected electricity demand and utilization as well as their capacity. A nameplate figure on its own does not show how much electricity a facility will actually use.
Global demand is rising, but estimates and forecasts are dated
The IEA’s 2025 Energy and AI report estimated that data centres used 415 TWh, about 1.5% of global electricity, in 2024. Its 2025 Base Case projected around 945 TWh of data-centre electricity use in 2030. In an April 2026 update, the IEA estimated 485 TWh in 2025 and about 950 TWh in 2030. These are estimates and projections from different publication dates, not a single timeless measurement.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstall#1 Best Overall
The IEA’s April 2026 update reported that data-centre electricity use rose 17% across the sector in 2025, while AI-focused centres’ use rose 50%. It projected AI-focused data-centre electricity use to triple from 2025 to 2030. These growth figures describe electricity use, not the growth rate of every facility or the share of all electricity attributable to AI.
AI is a major driver of expected growth, particularly through accelerated servers, but traditional servers and facility infrastructure also contribute. Cooling efficiency adds another source of variation: in its 2025 analysis, the IEA reported cooling at about 7% of electricity use in efficient hyperscale centres and over 30% in less-efficient enterprise centres. Those figures compare facility efficiency examples, not AI with traditional workloads directly.
Do AI data centres use more water?
Not necessarily. AI workload alone does not determine water demand, and the evidence here does not establish a robust, universal water-use ratio between AI-focused and traditional data centres. Cooling technology, local climate and the electricity supply mix all affect the result.
Rank #2
When assessing a particular site, distinguish water used directly for cooling from water associated with generating its electricity. Also check whether the reported figure is water withdrawn from a source or water consumed, what the source is, the time period covered and whether the local watershed is water-stressed. Without those details, a single water figure—or a “per AI prompt” figure—can give a misleading impression of local impact.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →What do data centres mean for local power grids and bills?
A modest global share of electricity use can still translate into a sizeable load in a particular place. The IEA’s 2025 analysis reported that nearly half of US data-centre capacity was concentrated in five regional clusters and noted substantial data-centre shares in some local electricity markets. Concentration is why a global percentage cannot answer whether a specific grid has room for a new facility.
A new data centre does not automatically raise local electricity bills, and it does not guarantee that bills will stay unchanged. The effect depends on local supply, when the facility uses power, grid headroom, required network investment and policy—including how costs are allocated. In a tight system, new demand may trigger investment; where existing infrastructure has spare supply, added demand may improve its use. Connection delays and the party responsible for paying for upgrades are practical local concerns.
Renewable-energy contracts are relevant, but they should not be treated as identical to the electricity physically serving a facility at every hour. For a specific project, ask what generation serves the load over time, what grid upgrades are needed, when the connection is expected and who pays for new generation or network costs.
How do emissions compare?
Emissions depend on the electricity generation serving the facility and the accounting boundary used. The IEA’s 2025 analysis estimated about 180 million tonnes (Mt) of indirect CO2 emissions from data-centre electricity use. That estimate covers data centres across workloads, with AI as a subset, and excludes emissions from backup power generation. It is not an AI-only figure or a measurement for any particular site.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesFor a site-level comparison, identify its location, the electricity mix used for the estimate and whether the calculation includes backup generation. Two facilities with similar electricity demand can have different electricity-related emissions if their power supplies differ.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What to compare when a data centre is proposed nearby
For a meaningful comparison between a proposed project and existing facilities, look for project-specific disclosures rather than relying on the labels “AI” or “traditional.” Useful details include:
- IT capacity and workload: what equipment is planned and whether the facility will support AI, other computing, or a mix.
- Expected utilization and electricity demand: anticipated usage over time, not capacity alone.
- Power supply: the generation mix physically serving the site, and how any renewable-energy arrangements relate to its hourly demand.
- Cooling and water: cooling design, water source, withdrawals versus consumption, reporting period and local water stress.
- Grid connection: available headroom, connection timing, anticipated upgrades and who is expected to pay for them.
- Emissions accounting: location, electricity-supply assumptions and whether backup generation is included.
The answers determine whether a particular facility is a substantial new burden locally; category-wide comparisons cannot settle that question.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Recommended Free Tools
