A neoscaler is a new type of cloud provider, often called a “neocloud,” that specializes in delivering high-performance infrastructure for intensive artificial intelligence (AI) workloads.

Unlike traditional hyperscalers (like Amazon, Google, and Microsoft), neoscalers focus specifically on providing GPU-as-a-service (GPUaaS) to meet the rapidly growing demand for AI computing.

Who qualifies as a neoscaler?

The term “neoscaler” covers a broad range of companies and services:

• AI developers: Companies that build and operate their own large-scale AI infrastructure, such as OpenAI.
• GPUaaS providers: Cloud providers that focus on renting out cutting-edge GPUs. Examples include CoreWeave, Lambda Labs, and Nebius.
• Smaller data centers: Smaller colocation or cloud providers that are capitalizing on the AI boom by building dedicated, high-capacity networks and GPU infrastructure.
• Legacy tech companies: Established companies like Oracle and Apple, which are building their own dedicated networks to support AI infrastructure.

Why are neoscalers on the rise?

The emergence of neoscalers is a direct result of the AI boom, as traditional cloud providers have struggled to keep up with demand for high-end GPUs. By specializing in AI, neoscalers offer a compelling alternative that provides:

• Access to scarce resources: Neoscalers give AI teams access to the most powerful GPUs without the long waiting periods often seen with hyperscalers.
• Cost-effectiveness: For AI-specific workloads, neoclouds can often offer more affordable prices.
• Better performance: The specialized, high-speed networks and purpose-built infrastructure of neoscalers are optimized for AI’s demanding performance requirements.

When did Neoclouds start to emerge in the US?

Neoclouds began emerging in the U.S. in the early 2020s, driven by the explosive demand for specialized graphics processing unit (GPU) power for artificial intelligence (AI). While some established companies shifted focus, the primary growth came from new, AI-native startups.

Key factors that contributed to the rise of neoclouds in the U.S. include:

• The generative AI boom (2023–2024): The widespread use of generative AI models led to an unprecedented demand for GPU computing power that traditional cloud providers struggled to meet.
• GPU scarcity: With traditional cloud companies unable to keep up, neoclouds quickly acquired GPUs and attracted significant funding to build AI-specific infrastructure.
• Repurposed resources: Former cryptocurrency miners, who had experience managing large GPU operations, adapted their infrastructure to serve AI compute needs.

Notable early players

• Crusoe Started cloud operations in the U.S. in mid-2022 with a small deployment in Montana before rapidly expanding to Virginia and Texas in late 2023.
• Lambda Labs Although founded in 2012, Lambda evolved into a prominent neocloud by offering specialized cloud and on-premises GPU solutions for AI researchers.
• CoreWeave Considered one of the largest neocloud companies today, CoreWeave saw its most significant growth in recent years as it scaled to meet AI demand.
• Vultr An older data center company, founded in 2014, that successfully pivoted its business to focus on AI compute in response to market demand.

What is Driving the use of NeoClouds?

• GPU Shortages: Neoclouds have emerged to provide much-needed access to high-performance GPUs, overcoming the multi-month waitlists and capacity constraints faced by traditional cloud providers.
• Cost-Effectiveness: These specialized providers offer more predictable pricing and cost reductions compared to hyperscalers, often by locating data centers with cheaper power resources.
• Performance: Neoclouds utilize bare-metal hardware, which avoids the virtualization penalties of hyperscalers, resulting in faster, more predictable throughput vital for AI workloads.

Market & Investment Momentum

• Rapid Adoption: A high percentage of organizations are already using or planning to adopt neoclouds for AI infrastructure, signaling a major shift in enterprise strategy.
• Venture Capital & IPOs: Investors have poured billions into the sector, as evidenced by significant capital raises for companies like Nebius and CoreWeave, whose stock prices have soared post-IPO.
• Hyperscaler Partnerships: Traditional cloud giants are partnering with and investing in neoclouds, even becoming customers of providers like CoreWeave and Nebius.

Operational & Strategic Growth

• Specialized Infrastructure: Providers are building out dedicated, high-performance bare-metal infrastructure optimized for AI, including custom ASICs and specialized GPUs.
• Regional & Hybrid Strategies: New regional players are emerging, offering local alternatives and catering to digital sovereignty needs. Hybrid models are also growing, combining neoclouds for training with hyperscalers for inference.
• AI-as-a-Service: A shift is occurring from just renting raw GPU power to offering pre-trained models and AI solutions, making AI projects easier and faster to launch.

CLICK SUBSCRIBE to ClearTech Loop on LinkedIn — for more field guides and conversations CIOs and CISOs can actually use. https://www.linkedin.com/newsletters/7346174860760416256/