As the demand for artificial intelligence (AI) and generative AI (GenAI) surges, Gartner warns that power consumption in data centres is set to explode, potentially outstripping utility providers’ capacity. By 2027, Gartner predicts that 40% of existing AI data centres will confront operational constraints due to insufficient power availability.
Gartner VP analyst Bob Johnson highlighted the situation's urgency: “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.” He emphasised that this impending power shortage could severely limit the establishment of new data centres and the scalability of GenAI applications.
Forecasts indicate that the energy required for AI data centres will soar to 500 terawatt-hours (TWh) annually by 2027—2.6 times the demand recorded in 2023. The rapid expansion of large language models (LLMs) necessitates extensive data processing capabilities, further exacerbating the power demand.
Gartner warns that the constraints on power supply will not only hinder the growth of new data centres but will also lead to higher electricity prices. “Significant power users are working with major producers to secure long-term guaranteed sources of power independent of other grid demands,” Johnson noted. This strategy reflects the urgent need for data centre operators to secure reliable energy sources amid rising costs.
The repercussions of this energy crisis extend beyond financial implications; sustainability goals are also at risk. The pressing demand for quick power solutions may lead to the continued operation of fossil fuel plants, hindering efforts to reduce CO2 emissions. Johnson stated, “Increased data centre use will lead to increased CO2 emissions to generate the needed power in the short-term,” complicating the journey toward zero-carbon targets.
Gartner advises organisations to reassess their energy strategies and sustainability objectives in light of the projected power shortages. Companies are encouraged to negotiate long-term power contracts and explore alternative strategies, such as edge computing, to mitigate energy demands.
The rapid growth of AI and GenAI is set to collide with energy supply limitations, forcing data centre operators and organisations to adapt swiftly to ensure sustainable and reliable operations in the coming years.