The economics and complexity of training generative AI models leave most enterprises with little choice but to rely on hyperscalers which can handle enormous workloads, from cloud computing to media streaming. They often use more than 5,000 servers and span millions of square feet. 

Yet large language models demand colossal computing power, built around graphics processing unit (GPU) clusters capable of handling the intense matrix calculations at the heart of neural networks. However, GPUs remain both scarce and staggeringly expensive, with the cost of building and operating dedicated clusters quickly reaching into the millions once power, cooling, networking and specialist staff are factored in. For the majority of firms, establishing such infrastructure in-house is prohibitively costly and operationally daunting.

The rise of Neoclouds 

Instead, a new generation of cloud providers has stepped into the gap, known as Neoclouds. These firms offer access to advanced GPU clusters and AI development environments through flexible, pay-as-you-go models. In essence, they allow businesses to rent high-performance computing power as needed, without the burden of owning or managing the hardware. This democratises access to AI capabilities, enabling smaller players to develop, train, and deploy large models that would otherwise be far beyond their reach.

Neoclouds are also reshaping the physical and commercial fabric of the data centre industry. Acting as a middle layer between hyperscalers and traditional colocation operators, they lease capacity in data centres across regional hubs, emerging markets, and major tier one cities. Their growth introduces new operational challenges, especially the race to energise sites quickly and secure affordable electricity at a time when power availability has become one of the sector’s greatest constraints.

Who are the protagonists? 

Among the leaders of this new wave is CoreWeave, whose IPO earlier this year cemented its status as the flagship of the Neocloud model. Europe, too, has strong market players, with Nscale and Nebius ranked second and third globally. Hardware manufacturer NVIDIA has become a pivotal force behind this ecosystem, holding stakes in four of the top five Neocloud firms and strengthening links between chip production and AI cloud infrastructure. Other fast-rising names, such as Vultr, Together AI, Voltage Park, and Fluidstack, signal that competition in this space is only just beginning.

Neoclouds are, in short, the agile, GPU-driven offspring of the cloud era, designed to power the next leap in AI innovation.

Further information

Contact Rupert Duckworth or Lydia Brissy

For further information, please read EMEA data centres – Between gridlock and gold rush