📊 Full opportunity report: The Neocloud Cartel: How the AI Industry Started Renting Compute From Itself on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
AI companies increasingly rent compute from each other, forming a cartel led by Nvidia. This shift decouples ownership from use and concentrates power among a few firms, raising questions about market fragility.
AI firms in 2026 no longer own most of their hardware; instead, they rent from a small circle of GPU landlords, with Nvidia at the center. This shift has created a tightly interconnected rental ecosystem that controls access to compute power, a critical resource for AI development.
The rise of the neocloud category — hyperscaler GPU-as-a-service providers — was driven by a GPU shortage in 2024–25, pushing companies to rent hardware instead of owning it. Major players like CoreWeave, Meta, and OpenAI have contracts worth tens of billions, all using Nvidia hardware.
In May 2026, xAI, a frontier AI lab, became a surprising landlord by leasing its supercomputer to competitors Anthropic and Google, paying over $26 billion annually. This marked a significant departure from traditional ownership models, emphasizing renting over owning hardware.
The financial flow reveals a circular pattern: firms like OpenAI have committed over $1 trillion in compute, with much of that money flowing back to Nvidia and other chip suppliers through investments, pre-purchases, and financing deals. Nvidia alone invests heavily in its customers and holds equity in many firms, effectively controlling access to the chips that power AI models.
Jensen Huang, Nvidia’s CEO, has indicated that a single gigawatt of AI data center capacity costs around $50 billion, with Nvidia capturing the majority of these funds through sales and investments. The allocation of GPUs, therefore, becomes a critical lever for market power, as Nvidia can decide who gets hardware in a supply-constrained environment.
Contracts often include governance clauses, such as xAI’s lease to Anthropic, which preserves Musk’s right to reclaim capacity if Anthropic’s AI harms humanity. This intertwining of supply, finance, and governance creates a fragile but powerful cartel, where access is gatekept and repriceable.
The Neocloud Cartel
Almost no one racing to build AI owns the machine it runs on. They rent — increasingly from each other — and the money loops back to one chip maker that’s also an investor in nearly everyone at the table.
The cartel isn’t a conspiracy — it’s the endpoint of extreme capital intensity, real scarcity, and one dominant supplier. But the same circularity that makes it powerful makes it a fuse: each cancelled order is someone else’s missing revenue. Don’t be a price-taker at the bottom of a loop you don’t control — own your inference, keep an open-weight fallback, diversify silicon.
Implications of the AI Compute Cartel for Industry Power
This evolving compute cartel concentrates power among a few firms, especially Nvidia, which controls the supply chain and financing. Such a structure raises concerns about market dominance, potential bottlenecks, and the fragility of AI development if supply chains are disrupted or if the cartel’s internal agreements break down. It also signals that ownership of hardware is increasingly decoupled from AI research and deployment, altering competitive dynamics.
Local AI on Linux in Practice: Build Private LLM Servers, GPU Workstations, Ollama Apps, Dockerized AI Services, and Self-Hosted AI Infrastructure with CUDA, ROCm, vLLM, and Open WebUI
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Formation of the Neocloud and the Rise of Compute Renting
Since 2024, the AI industry has faced a severe GPU shortage, leading firms to rent hardware instead of owning it. The emergence of the neocloud category — specialized hyperscaler GPU providers — has reshaped the supply landscape. Major investments from Meta, OpenAI, and others, along with the rise of companies like CoreWeave, have created a market where hardware is largely leased, not owned.
In 2026, the pattern shifted further as xAI leased its supercomputer to competitors, highlighting a move toward a rental-based ecosystem. This dynamic is driven by the high costs and scarcity of Nvidia chips, with Nvidia acting as the central gatekeeper, holding significant equity stakes and financing arrangements with key players.
“A gigawatt of AI data center capacity costs roughly $50 billion, and Nvidia captures the majority of those dollars.”
— Jensen Huang, Nvidia CEO
AI hardware rental platforms
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unclear Risks and Potential Instability of the Cartel
It remains uncertain how sustainable this tightly interconnected rental ecosystem is, especially if supply chains are disrupted or if Nvidia’s control is challenged. The fragility of the cartel could lead to market shocks or shifts in power, but specifics are still emerging.
NVIDIA RTX PRO 6000 Blackwell Server Edition
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Developments in AI Hardware Supply and Market Dynamics
Expect increased scrutiny of Nvidia’s market power and potential regulatory responses. Additionally, more firms may seek alternative supply sources or develop proprietary hardware to reduce dependency. Monitoring how the rental ecosystem evolves and whether new entrants can challenge the current cartel will be key.
hyperscaler GPU as a service
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why are AI companies renting hardware instead of owning it?
Due to a GPU shortage in 2024–25, renting became the only practical way for many firms to access the necessary compute power quickly without building their own infrastructure.
What role does Nvidia play in this ecosystem?
Nvidia is the dominant chip supplier, controlling the majority of GPU supply, investing heavily in its customers, and holding equity in many firms, effectively acting as the gatekeeper of AI compute access.
Could this rental cartel lead to market instability?
Yes, the high interdependence and control by a few firms create a fragile system that could face disruptions if supply chains break or if Nvidia’s dominance is challenged.
What are the implications for AI development if ownership is decoupled from use?
It could centralize power in the hands of a few firms, potentially stifle competition, and make the industry more vulnerable to supply disruptions or strategic conflicts.
Source: ThorstenMeyerAI.com
