The revenue generated by neocloud service providers, who are seemingly usurping the traditional hyperscale providers when it comes to hosting artificial intelligence (AI) workloads, is expected to exceed $23bn during 2025.

That’s according to market data shared by Synergy Research Group (SRG), which confirmed the revenue generated by neoclouds surpassed $5bn during the second quarter of 2025, which is up 205% on the previous year.

Throughout the course of 2025, there have been numerous announcements about neocloud providers, such as CoreWeave and the OpenAI-backed Stargate project, announcing the opening or future development of datacentre sites that will be used to house compute-intensive AI workloads.

The purveyors of neocloud services specialise primarily in the provision of the graphics processing units (GPU) infrastructure needed to run AI workloads, and – for this reason – are sometimes referred to as being GPU-as-a-service (GPUaaS) providers.

This is in contrast to the more traditional hyperscale providers who offer a broader range of cloud services, rather than just GPU infrastructure services.  

The uptick in demand for neocloud services has, in turn, given rise to a “lot of hype and exaggerated or unsubstantiated claims about the massive buildout of gigawatt campuses” to underpin the projected growth of generative AI workloads, said SRG in its research note.

However, it predicts that the revenue generated by neoclouds will hit $180bn by 2030, albeit with “road bumps” along the way, in the form of financing difficulties, power supply issues and problems sourcing the necessary hardware.

However, it is SRG’s view that the underlying demand for AI services will help the neocloud sector overcome these barriers and other competitive pressures, on account of the fact that providers are targeting the fast-growing AI segment.

“Neocloud plans and initiatives remain somewhat fluid, but in aggregate, Synergy sees them steadily gaining market share in those high-growth target market segments,” said SRG in its research note.

“In terms of being a direct competitor to the traditional hyperscale cloud providers, CoreWeave is leading the group. While its focus and business model is different from the other neoclouds, OpenAI is the largest of the group, and will remain a key player over the coming years thanks to its Stargate initiatives.”

As previously reported by Computer Weekly, the OpenAI-backed Stargate initiative has seen the ChatGPT maker embark on joint ventures with the likes of GPU provider Nvidia and neocloud firm Nscale in the UK, in the interests of providing sovereign compute capacity for the sole purpose of hosting AI models.

Typically, neoclouds are either relatively new startup companies or, as is the case with Nscale, are crypto mining companies that have pivoted to providing high-performance computing services, said SRG.

“There is a lot of hype around neocloud, Stargate and gigawatt campus developments, but when you work through the marketing smoke and mirrors and look at the underlying numbers, the growth rates and future market size are truly impressive,” said SRG founder and chief analyst Jeremy Duke.

“GPUaaS and GenAI platform services are currently growing at around 165% per year and neoclouds are gaining share in those high-growth markets,” he added. “There is every reason to believe that they will continue to grow their market share, as cloud providers of all types struggle to match supply with burgeoning AI demand.”

Read more on Infrastructure-as-a-Service (IaaS)

• 
  

  Zayo, Equinix develop AI network framework

  By: Joe O’Halloran

• 
  

  Nscale explained: Everything you need to know

  By: Caroline Donnelly

• 
  

  AI infrastructure provider Nscale secures $1.1bn in series B funding

  By: Caroline Donnelly

• 
  

  IT leaders face data infrastructure gaps as AI workloads grow

  By: Simon Robinson