Microsoft Cancels UK Data Centre Deal as Nscale Faces Power Supply Delays- Microsoft has withdrawn from an agreement to lease computing capacity at a planned data centre in Essex, dealing a setback to British artificial intelligence infrastructure start-up Nscale as delays in securing electricity threaten its expansion plans.
The US technology company has cancelled its agreement covering Nscale’s proposed facility in Loughton after it became clear that the site was unlikely to receive a grid connection until after 2030, according to a person familiar with the matter cited by the Financial Times.
The development highlights a growing obstacle for the AI industry: securing enough electricity to power the vast data centres required to support increasingly demanding artificial intelligence systems.
The Essex facility had been expected to connect to the electricity grid this year. However, that timetable is no longer considered achievable, undermining the project’s ability to deliver computing capacity within the timeframe expected by major technology customers.
For Microsoft, the decision means stepping away from a planned source of computing infrastructure. For Nscale, it raises questions about how quickly the company can bring its projects online as it pursues ambitious growth plans.
Electricity Shortages Threaten AI Expansion
The cancellation comes as technology companies race to build the infrastructure needed to train and operate advanced AI models. These systems require enormous computing resources, including specialised processors, high-speed networking equipment and sophisticated cooling systems.
Yet access to electricity is increasingly becoming a limiting factor in the development of new facilities.
Data centre operators can secure financing, purchase hardware and prepare sites, but those investments cannot deliver their intended value without a reliable power supply. Connecting large facilities to the electricity grid can involve complex planning, infrastructure upgrades and lengthy waiting periods.
The situation creates a significant challenge for companies seeking to expand rapidly in response to demand for AI services. Even when a project is technically ready, delays in power connections can prevent operators from supplying computing capacity when customers need it.
Microsoft’s withdrawal from the Loughton agreement illustrates how infrastructure constraints can influence commercial decisions across the sector. For customers investing heavily in AI, the timing of capacity delivery is critical, making projects with uncertain electricity connections less attractive.
Major Setback for Nvidia-Backed Nscale
Nscale has emerged as a prominent player in the growing market for AI-focused cloud infrastructure, commonly described as the neocloud sector. Such businesses provide specialised computing resources, often built around high-performance Nvidia processors, to companies developing and deploying AI applications.
The company is backed by Nvidia and has attracted attention for its ambitions to build infrastructure capable of serving the rapidly expanding AI market.
Its board includes former Meta executives Sir Nick Clegg and Sheryl Sandberg, adding to the company’s high-profile leadership connections.
According to the Financial Times report, Nscale is targeting a New York stock market listing this year at a valuation of $35 billion. The loss of Microsoft’s agreement at the Essex site represents an unwelcome development as it seeks to establish itself as a major provider of AI computing capacity.
A proposed valuation of that scale places considerable importance on the company’s ability to demonstrate that its infrastructure investments can translate into operational capacity and commercial revenue.
The cancellation does not, by itself, establish how significantly Nscale’s overall business or potential listing will be affected. The company may have other projects and customers, and the available information does not indicate that its broader expansion strategy has been abandoned.
Nevertheless, the Essex setback underlines the risks associated with building an AI infrastructure business in a market where demand can grow faster than essential utilities can be delivered.
Microsoft’s AI Infrastructure Strategy Faces a Practical Constraint
Microsoft has committed substantial resources to expanding its AI capabilities and cloud computing operations. Access to dependable computing capacity is essential to supporting its products, enterprise customers and partnerships across the AI sector.
Withdrawing from a project facing a lengthy power delay may therefore reflect the importance of securing infrastructure that can become operational within commercially useful timeframes.
The decision also illustrates a broader tension within the technology industry. Companies are racing to acquire advanced chips and develop new facilities, but access to those components is only one part of the equation. Electricity, land, cooling systems and grid infrastructure must also be available.
The Loughton project demonstrates how a delay in a single critical resource can disrupt an otherwise ambitious development plan.
A Wider Warning for the AI Industry
The power challenge is becoming increasingly important as governments and technology companies seek to expand AI infrastructure. Data centres consume substantial amounts of electricity, and the arrival of larger computing facilities can require major investment in local energy networks.
Where grid capacity is limited, new projects may face long waits for connections, forcing operators to reconsider schedules, locations or commercial agreements.
For Britain, such delays could complicate efforts to attract investment in AI infrastructure and establish itself as a competitive destination for advanced computing.
The immediate consequences of Microsoft’s decision will depend on how Nscale responds to the setback and whether alternative arrangements can be made for the affected project.
For the wider industry, however, the message is clear: access to advanced chips and financial backing will not be enough to guarantee rapid AI expansion. Without sufficient electricity and the infrastructure needed to deliver it, even heavily backed projects can struggle to meet their deadlines.
As competition for AI computing capacity intensifies, reliable power supplies could become just as important as processors in determining which companies can expand fastest.
