📊 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.
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
AI companies are increasingly renting compute from each other, creating a closed loop centered on Nvidia. This shift has transformed the industry into a cartel with concentrated power, raising questions about stability and control.
In 2026, major AI firms are no longer owning the hardware they run on; instead, they are renting compute from a tightly interconnected group of companies, with Nvidia at the center. This shift has created a de facto cartel controlling access to critical AI infrastructure, with significant implications for industry power dynamics.
Recent reports reveal that AI companies such as xAI, Anthropic, and Google are leasing large-scale supercomputers from each other, often paying hundreds of millions to over a billion dollars monthly. Nvidia dominates as the primary supplier, with its chips and infrastructure forming the backbone of this ecosystem. In particular, xAI’s lease of its Colossus 1 supercomputer to Anthropic and Google exemplifies how ownership has decoupled from use, turning compute into a rented resource.
Financial arrangements reveal a circular flow of capital: Nvidia has invested heavily in AI firms like OpenAI and Anthropic, financing their hardware needs and holding equity stakes. The industry’s largest players are now intertwined through multi-billion-dollar agreements, with Nvidia capturing the majority of the revenue and controlling chip allocations. This concentration of control effectively makes Nvidia the gatekeeper of AI compute resources, giving it immense influence over the industry’s future.
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 development signifies a fundamental shift in how AI infrastructure is controlled, moving from ownership to leasing within a tightly knit network. The concentration of power in Nvidia and a few firms creates a fragile but highly influential cartel that could determine which companies succeed or fail based on access to compute resources. This structure raises concerns about competition, innovation, and the potential for supply chain disruptions that could impact the entire AI industry.
As an affiliate, we earn on qualifying purchases.
Background of the Growing AI Compute Market and Its Players
The AI industry has faced a severe GPU shortage since 2024-25, prompting companies to rent hardware instead of owning it. CoreWeave, Meta, OpenAI, and others have all relied on Nvidia’s chips, creating a market where access is limited by supply and controlled by a small number of suppliers. The emergence of ‘neocloud’ hyperscalers like CoreWeave, backed by billions in contracts, exemplifies this shift. In May 2026, xAI’s decision to lease its supercomputer to competitors marked a turning point, illustrating how ownership of hardware has become secondary to leasing arrangements.
Prior to this, the industry was characterized by competitive hardware procurement; now, it revolves around a circular financing loop where a handful of firms finance each other’s compute needs, reinforcing Nvidia’s central role.
“A gigawatt of AI data center capacity costs roughly $50 billion, with most of that flowing to Nvidia.”
— Jensen Huang, Nvidia CEO
enterprise AI compute rental services
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Uncertainties Surrounding the Stability of the Compute Cartel
It is not yet clear how fragile this cartel is or whether it can sustain itself amid potential supply disruptions or regulatory scrutiny. The dependence of firms on Nvidia and each other creates vulnerabilities that could lead to fractures, but the full extent of these risks remains uncertain as the industry continues to evolve.
high performance GPU for AI training
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for the AI Compute Ecosystem and Industry Regulation
Industry observers expect increased scrutiny from regulators regarding anti-competitive practices as the concentration of compute resources consolidates. Additionally, companies may seek alternative supply chains or develop in-house hardware to reduce dependence. The coming months will likely reveal whether this cartel can maintain its cohesion or if cracks will appear under pressure.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the ‘Neocloud Cartel’?
The ‘Neocloud Cartel’ refers to the small, interconnected group of AI companies and suppliers that rent compute from each other, primarily controlled by Nvidia, creating a concentrated power structure in AI infrastructure.
Why is Nvidia so central to this system?
Nvidia supplies the majority of the chips used in AI data centers, controls chip allocation, and has invested heavily in AI firms, making it the key gatekeeper of AI compute resources.
What risks does this cartel pose to the industry?
The cartel’s concentration could lead to supply bottlenecks, reduced competition, and vulnerability to regulatory actions, potentially disrupting AI development and deployment.
Could companies develop their own hardware to bypass this system?
While some firms might attempt in-house hardware development, the high costs and complexity make it unlikely to fully circumvent the existing supply chain in the near term.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
![Rob Pike – 'Concurrency Is Not Parallelism' [video]](https://deepintellica.com/wp-content/uploads/2026/07/rob-pike-concurrency-is-not-parallelism-video-featured-560x300.jpg)
