Global data centre power demand will reach 132 gigawatts in 2026, according to Gartner. In Australia, demand is expected to rise to almost 1.5 gigawatts in the same year.
That would represent annual growth of 27% worldwide from 104 gigawatts in 2025, while the Australian market is forecast to expand 38.3% from 1.1 gigawatts.
By 2030, Gartner expects global data centre power demand to climb to 290 gigawatts. In Australia, it forecasts demand will reach almost 3.8 gigawatts by the end of the decade.
Electricity consumption is also set to rise sharply. Worldwide data centre electricity use is projected to increase 26% in 2026 to 565 terawatt hours from 447 terawatt hours a year earlier.
In Australia, data centre electricity consumption is expected to rise 37.7% in 2026 to 6.2 terawatt hours from 4.5 terawatt hours in 2025. By 2030, it is forecast to reach 15.7 terawatt hours.
Growth is being driven by demand for artificial intelligence workloads, particularly the spread of AI-optimised servers, which consume large amounts of electricity and add pressure to power networks and cooling systems.
AI-optimised server adoption will account for 31% of global data centre power consumption in 2026, Gartner said. In Australia, the share is expected to be higher at 35.7%, equal to 2.2 terawatt hours.
By 2027, electricity consumption from AI-optimised servers is expected to overtake that of conventional servers both globally and in Australia, underlining how quickly computing demand is shifting inside data centres.
Worldwide, conventional servers are forecast to consume 195 terawatt hours in 2026, up only slightly from 193 terawatt hours in 2025. AI-optimised servers, by contrast, are projected to use 175 terawatt hours in 2026, up from 95 terawatt hours a year earlier.
Cooling and other infrastructure are also taking a larger share of electricity use. Gartner estimates those systems will consume 195 terawatt hours globally in 2026, up from 159 terawatt hours in 2025.
Power pressure
Rapidly rising electricity demand is creating a supply constraint for the sector. Gartner estimates data centre electricity consumption will exceed 1,200 terawatt hours by 2030, a level that would leave grid supply unable to support all planned new construction.
That could affect operators, cloud providers and corporate users seeking space in new facilities, particularly in markets where grid connections are already hard to secure or utilities face long lead times for upgrades.
"Surging demand for compute-intensive AI workloads is driving unprecedented data centre power growth, while AI capacity is now constrained by power availability, making data centre power security the new battleground for scaling and protecting margins in the global AI race," said Linglan Wang, Director Analyst, Gartner.
Infrastructure and operations leaders will need to respond by focusing on energy efficiency as well as access to electricity, Gartner said. It pointed to cooling systems and edge computing as areas where operators may try to reduce pressure on centralised facilities.
"Infrastructure and operations (I&O) leaders must prioritise efficiency upgrades and secure grid access," said Wang. "They also need to invest in high-efficiency cooling systems and edge computing to mitigate power constraints and ensure sustainable, scalable growth."
The forecasts suggest Australia may face a sharper near-term rise than the global average, reflecting the pace of local investment in AI infrastructure and the relatively small existing base of data centre power demand.
With AI-optimised servers set to account for more than a third of Australian data centre electricity consumption in 2026, operators in the country may face earlier pressure to expand power access and manage rising energy use within existing sites.
Globally, Gartner's figures show total data centre electricity consumption rising from 447 terawatt hours in 2025 to 565 terawatt hours in 2026 and 702 terawatt hours in 2027, with AI-optimised systems and cooling infrastructure accounting for most of the increase.