AI and generative AI (genAI) are triggering a rapid rise in electricity consumption, with data centres set to reach 160 per cent growth over the next two years, according to new research from Gartner.

The technology research firm said with this growing consumption, around 40 per cent of existing data centres will be operationally restricted by power availability by 2027.

Gartner estimates the power required for data centres to run incremental AI-optimised servers will reach 500 terawatt-hours (TWh) per year in 2027, which is 2.6 times the level in 2023.
As a result, Gartner predicts that electricity prices will increase and there could be power shortages.

This in turn will lead to higher operating costs for systems using large language models (LLMs) and the number of new data centres will be determined by the availability of power to run them, the company warned.

Gartner is urging organisations to determine the risks potential power shortages will have on all products and services and has recommended having a plan in place to negotiate long-term contracts for power.

Organisations will need to factor in “significant” cost increases when developing plans for new products and services, with Gartner advising firms to look for alternative approaches which require less power.

“The explosive growth of new hyperscale data centres to implement genAI is creating an insatiable demand for power that will exceed the ability of utility providers to expand their capacity fast enough,” said Bob Johnson, VP analyst at Gartner. “In turn, this threatens to disrupt energy availability and lead to shortages, which will limit the growth of new data centres for genAI and other uses from 2026.”

Elsewhere, Gartner also said that sustainability goals will be affected by solutions to find more power, as suppliers increase production. This could lead to phasing out fossil fuel plants later than planned to keep up with demand.

Gartner said firms may need to re-evaluate their sustainability goals relating to CO2 emissions due to data centre requirements.

Additionally, the company explained that when developing genAI applications, they should focus on using a minimum amount of computing power and look for alternatives such as smaller LLMs.

---