Microsoft Exits Delayed Essex Data Centre Deal as Power Access Slips to 2030s

Microsoft has withdrawn as anchor tenant for Nscale’s planned £2-billion data centre in Loughton, Essex, after the site’s expected power connection was pushed to the 2030s. The move leaves the site without an announced customer and comes as Nscale pursues a possible flotation of up to $35 billion (£26 billion).

Slashdot has reported:

Microsoft has pulled out of a planned data centre deal in Loughton, Essex, after delays to the site’s electricity connection pushed its expected power supply into the 2030s. The facility, being built by Nvidia-backed Nscale, was planned as a £2-billion project, with Microsoft signed up as its anchor tenant.

The data centre had been due to receive power at the end of the year, but Nscale was told the connection would not arrive until the 2030s. Microsoft had agreed to rent 23,000 Nvidia AI chips at the facility and was its only announced customer, leaving the site in apparent limbo after its withdrawal.

Nscale and Microsoft continue to work together on other projects in Norway, Portugal and America. The Essex exit threatens to undermine Nscale’s effort to pursue a Wall Street flotation valued at up to $35 billion (£26 billion), despite a sharp sell-off in AI computing stocks.

The delay reflects a wider constraint on British data centre development: grid connection times have become the biggest brake on new facilities. Some UK projects are waiting close to a decade for enough power, and one estimate found that more than a quarter of data centre capacity was delayed in 2025 alone.

The withdrawal comes amid investor concerns about whether data centre developers will be able to repay their debts as interest rates and inflation continue to rise. Rival developers’ share prices fell over the past week. Gartner said last week that $3.5 trillion could be written off by 2030 if the AI boom does not deliver on its promises.

Gartner said its research indicates that “enterprise and consumer AI revenue will be insufficient to support profitable returns on all planned AI capital expenditures during this period.” It expects AI providers to spend nearly $1 trillion on AI infrastructure in 2026 and invest a further $6 trillion from 2027 to 2030.

“Failure to achieve sustainable return on investment capital (ROIC) on rapidly depreciating assets would almost certainly necessitate write-downs,” Gartner wrote. Based on its projections, continuing investment in AI infrastructure could expose more than $4 trillion in asset value to write-down risk through 2032. Gartner does not expect a 12% ROIC to be reached before the end of 2032 in that scenario.

Gartner estimates spending on AI-optimized infrastructure as a service (IaaS) will rise from $66 billion in 2027 to $174 billion in 2030. “The growth in spending on AI-optimized IaaS will be too low for most AI compute providers to recover their AI investments before the current generation of AI semiconductors is replaced,” said Will Sommer, senior director analyst at Gartner. By 2030, Gartner predicts cost pressures and a lack of a differentiating business model will force more than 80% of neocloud providers, which specialize in AI infrastructure, to cease independent operations.