Neoclouds are no longer just a stopgap solution for GPU-starved AI companies; they're maturing into legitimate competitors that could reshape how enterprises buy computing power. These specialized cloud providers, which emerged in 2023 and 2024 when traditional hyperscalers couldn't meet demand, are now forcing Amazon, Microsoft, and Google to reckon with pricing, performance, and customer service in ways they haven't had to before.

What Exactly Is a Neocloud, and Why Should You Care?

A neocloud is a specialized cloud infrastructure provider focused on high-performance computing, artificial intelligence, and machine learning workloads. Unlike Amazon Web Services or Microsoft Azure, which offer everything from databases to analytics tools, neoclouds concentrate narrowly on delivering raw GPU computing power through GPU-as-a-Service and bare-metal infrastructure. The market was valued at $35.22 billion in January 2026 and is projected to reach nearly $240 billion within five years, according to industry research.

The origin story is unexpected. When cryptocurrency crashed in late 2022, crypto mining operations were left with racks of idle GPUs. Rather than scrapping the hardware, these infrastructure providers pivoted to the booming artificial intelligence market. "These people had the capability to run giant data centers built for high-performance computing," explained Paul Updike, senior director of technical marketing engineering at Nutanix. "It turns out, the infrastructure and hardware engineering skills you need for crypto are the exact same infrastructure and hardware engineering skills you need for AI".

Are Neoclouds Just a Temporary Trend, or Here to Stay?

Skeptics argue that neoclouds are merely filling a supply chain gap that will close as GPU availability normalizes and traditional hyperscalers catch up. Steve McDowell, Principal Analyst at NAND Research, compared current neoclouds to early colocation providers before cloud computing took off. "The neoclouds today look a whole lot more like an advanced Rackspace," he noted, highlighting structural vulnerabilities in the current model.

However, other analysts see a genuine future for these providers. Tech analyst David Linthicum argued in April 2026 that "hyperscalers are at real risk of losing market share as more enterprises switch to lower-cost, specialized competitors." He added that without meaningful price reductions on AI-specific services, "AWS, Microsoft, and Google face the prospect of ceding entire segments of the market to their smaller rivals".

Matt Kimball, principal datacenter analyst at Moor Insights and Strategy, initially dismissed neoclouds but changed his view. "I see the path," he said. "They built a very specific infrastructure for a very specific set of customers and use cases, and it works really well for them. And if you believe that inference is going to be the size that it's going to be, they're really well prepared to do that better than the major cloud service providers today".

How Neoclouds Must Evolve to Survive Long-Term

The key to neocloud survival lies in a fundamental shift from training AI models to running them in production, a phase called inference. Currently, neoclouds operate like beach house rentals, where a single customer runs a large training job across dedicated infrastructure. Production inference requires something closer to an apartment building, with multiple customers sharing the same infrastructure simultaneously.

• Multi-Tenancy Support: Neoclouds must develop true multi-tenancy capabilities, allowing multiple customers to safely share the same hardware without interference or security breaches.
• Enterprise-Grade Software Infrastructure: Moving beyond bare-metal offerings, neoclouds need sophisticated software platforms to manage workload scheduling, resource allocation, and performance guarantees across diverse customer needs.
• Production Inference Optimization: As inference workloads become dominant, neoclouds must optimize for lower resource consumption per job and the ability to burst capacity up and down dynamically rather than maintaining full clusters for single customers.

Updike emphasized the scale of this challenge: "In the world of inference, where you're actually running the models and doing the work, you're now in a space where you're not going to consume as many nodes. You may even burst into nodes and come back down out of nodes. It's a lot of jobs running simultaneously that you have to manage and schedule and make sure that they're going to play nice together".

Updike

McDowell suggested that Nutanix, a hybrid cloud infrastructure company, is well-positioned to help neoclouds make this transition. "Neoclouds have a vast number of resources, and they are going to need a control plane that helps manage workloads across those resources. That's what Nutanix excels at," he noted.

The Market Consolidation Ahead

While neoclouds currently number around 200 tracked providers, Kimball expects significant consolidation. "Five years from now, there aren't going to be 200 neoclouds. You're going to see a lot of these fade away," he said. Linthicum added that hyperscalers will likely acquire promising neoclouds, though he acknowledged these new competitors have already forced "a long-overdue reckoning on price, performance, and customer service" while giving customers more choices.

Linthicum

Beyond cost advantages, neoclouds may also appeal to enterprises with data sovereignty concerns. "Localized neoclouds provide a level of sovereignty that you don't have with the hyperscalers," Updike explained, suggesting that geographic data residency requirements could sustain neocloud demand even as the broader market matures.

The neocloud story represents a broader shift in how enterprises access AI infrastructure. What began as a temporary workaround for GPU scarcity has evolved into a genuine competitive force, one that's pushing the industry toward better pricing, specialized services, and customer-centric innovation. Whether neoclouds survive as independent entities or get absorbed into larger platforms, their impact on the AI infrastructure market is already undeniable.