
Once a spinoff of Arkon Energy—a firm tied to crypto-mining—Nscale has leapt into the big leagues. The UK-based startup recently landed a $700 million investment from Nvidia, Microsoft, and OpenAI to build its hyperscale AI data-center infrastructure. The plan calls for deploying tens of thousands of Nvidia Blackwell GPUs across new facilities, beginning with a major supercomputer campus in Loughton. This marks a shift in focus: from pure crypto hashpower toward providing cutting-edge compute power for AI researchers, enterprises, and sovereign workloads.
The move is part of a broader trend. As AI workloads demand exponentially more compute, companies with access to both capital and infrastructure are racing to outscale their peers. Nscale’s founders are betting that their experience in maintaining energy-intensive operations from their crypto roots gives them a unique advantage: an ability to optimize power usage, build out infrastructure in power-rich locales, and manage cooling and electrical supply at scale. With capacity initially pegged at ~50 megawatts (scalable to ~90 MW) and plans for 23,000 or more GPUs in the early stages, Nscale isn’t just building a data center—it’s constructing a platform for AI growth in the UK and potentially globally.
That said, the stakes are high and the risks considerable. Building at this scale means navigating regulatory hurdles, securing stable power deals, managing supply chain constraints for high-end hardware, and ensuring continued demand from AI applications. Energy cost volatility and environmental scrutiny pose additional challenges. If Nscale can deliver efficiency gains, strong utilization rates, and a path to profitability, it could become a major node in the global AI infrastructure network—providing an interesting contrast to traditional crypto miners, whose fortunes often depend heavily on Bitcoin price swings and mining difficulty. But success will depend on execution, as the difference between ambition and impact grows sharper in this AI arms race.