Demand for artificial intelligence (AI) data centres has been at an all-time high lately, with demand expected to exceed current capacity by 2027.
According to Salute's State of the Industry 2026 report, as cited by Data Centre News, 83 per cent of senior IT and data centre leaders expect AI workloads to exceed current capacity. Electricity availability seems to be the primary barrier to this expansion.
The report highlighted a critical need for advanced liquid cooling to support higher-density racks, as nearly 75 per cent of operators feel unprepared for the incoming AI scale.
Earlier today, Reuters reported that AI data centres' skyrocketing demand is exacerbating shortages of critical grid equipment such as transformers across the U.S. This is driving up costs, extending wait times as well as impelling utilities and developers to confirm purchase orders way ahead of time.
The soaring costs are causing power companies to hastily secure major grid equipment, including transformers, switchgear, and circuit breakers.
This has created severe shortages in the supply chain, so much so that the lead time to acquire equipment is up to 160 weeks for some resources.
AI Data Centre Demand Hampering Supply Chain
U.S. is particularly facing issues with limited electricity grids that are blocking the accelerating growth of AI infrastructure.
Large transformers are required to scale this rapid development as these transformers function as they adjust voltage levels to make long-distance power transmission efficient and safe. Unfortunately, the transformers are in short supply, impacting the entire AI data centre supply chain.
Experts told Reuters that supplies are being further squeezed by the rapid buildout of AI infrastructure. Power companies have had to not only lock in timelines but extend lead times for high-voltage transformers. Some suppliers have extended delivery time from about one year in 2020–2021 to several years today.
Ben Boucher, senior analyst with consultancy Wood Mackenzie, stated: "Equipment availability is becoming the biggest concern for developers as they value time to market so highly.”
He further noted that while large power transformers are experiencing the most pronounced market shortage, data centre construction is driving demand for equipment such as circuit breakers and switchgear, which are set to face larger market deficits.
Also Read: Top 10 AI Data Centre Companies for 2025
Transformer Shortage Sparked by COVID-19 Lockdowns
The transformer shortage is not sudden but has been observed over the last five years. It coincided with the time when AI demand began to rise. More importantly, the root of the issue can be traced back to the COVID-19 lockdowns, which disrupted production of transformers when demand recovered faster than supply.
There weren't enough transformers to meet everyone's needs.
Back in 2019, Greg McCulloch, now former CEO of Aegis Data, told EM360Tech that there’s no escaping the reality that as more connected devices and technology trends sweep the market, more demands will be placed on the data centre to provide the high-powered servers and cooling systems required.
“But in the pursuit of guaranteeing performance, it is having an accumulative effect on the global share of data centre emissions, which has been criticised for being on par with aviation."
In an episode of the Don’t Panic It’s Just Data podcast, EM360Tech’s host and Journalist, Shubhangi Dua, spoke to Rolf Bienert, Technical & Managing Director of the OpenADR Alliance, about the urgent need for sustainable energy practices within the data centres industry.
Reflecting on the processing power of AI, Bienert told Dua that AI needs a lot more processing power; a factor of 10 to 100 times the processing power as a Google search, if you have a simple AI search.
“We have all these resources, and on the other end we have the big consumers,” He added that the trick is to really aggregate the various energy generators and storage assets.
He suggested combining the power strategy with multiple energy resources, including Solar energy, battery systems and other generators, renewables (such as hydro) and potentially small-scale nuclear power plants in the future.
“It's really important now that the IT managers, CIOs and C-suites really actively try to gain an understanding about what's going on in the grid,” stated the technical & managing director.
Alluding to the standards of scaling energy solutions, Bienert says that to make it in the market, enterprises need to adhere to industry-wide standards for the energy transition to be economically viable and functional. This would allow vendors to plug new technology into the grid as easily as plugging a device into a wall outlet.
Comments ( 0 )