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 firms increasingly rent compute from each other, forming a tightly interconnected cartel dominated by Nvidia. This shift impacts industry power dynamics and introduces new vulnerabilities.
In 2026, the AI industry is operating within a tightly interconnected network of companies that rent compute from each other, rather than owning their own hardware. This shift, driven by a GPU shortage and massive investments, has created what analysts describe as a ‘neocloud’ cartel, with Nvidia at its core. This development fundamentally alters the power structure in AI infrastructure, as control over chip allocation and financing now determines industry influence.
Almost no AI companies own the hardware they run on; instead, they lease from a small group of GPU landlords, including Nvidia, CoreWeave, and others. Notably, xAI, a frontier AI lab, has leased its supercomputer to rivals like Anthropic and Google, signaling a move towards self-ownership of compute resources—yet still within a tightly controlled financial and contractual framework.
Major players such as OpenAI, Meta, and others have committed hundreds of billions of dollars to rent compute, with a significant portion of this money flowing back to Nvidia, which supplies the chips and holds stakes in many firms. Nvidia alone captures the majority of the industry’s compute spending, controlling chip allocation and thus industry access. This creates a power dynamic where Nvidia effectively governs who can compete in AI development.
The Neocloud Cartel — The Control Series, Part 2: Compute
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.
invests ~$100B
commits ~$1.15T
buy GPUs
+ equity stakes
NVIDIA
the chokepoint
THE LABS
OpenAI · Anthropic
CLOUDS & CHIPS
CoreWeave·Oracle·AMD
↻ each deal lifts
the next one’s value
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 emerging ‘neocloud’ cartel concentrates control of AI compute resources in the hands of a few firms, primarily Nvidia. Such concentration grants Nvidia and its allies significant influence over who can develop and deploy advanced AI models, potentially shaping the competitive landscape and innovation pace. However, this interdependent system also introduces fragility, as the entire structure relies on circular financing and supply agreements that could unravel if key relationships falter.
Formation and Evolution of the AI Compute Cartel
Over the past three years, a GPU shortage and rapid AI growth prompted companies to rent hardware rather than build their own. CoreWeave emerged as a major hyperscaler, and in 2026, xAI became a surprising participant by leasing its supercomputer to rivals, signaling a shift towards self-ownership within a rent-based system. The circle of financing involves giants like Nvidia, Microsoft, and Amazon, with billions of dollars flowing among them, creating a closed loop of dependency and control.
This network resembles a cartel more than a free market, with Nvidia at the center, controlling chip supply, and influencing pricing and access through contractual and financial leverage. The industry’s reliance on this small set of firms makes the entire ecosystem vulnerable to disruptions in supply, finance, or contractual agreements.
“A gigawatt of AI data center capacity costs roughly $50 billion, with about $35 billion flowing directly to Nvidia.”
— Jensen Huang, Nvidia CEO
Unclear Risks and Potential Disruptions to the Cartel
While the structure appears stable, it remains uncertain how fragile this cartel is in practice. The reliance on circular financing and contractual dependencies could lead to vulnerabilities if key relationships break down or supply constraints worsen. The impact of regulatory scrutiny or geopolitical restrictions on chip exports also remains unclear and could reshape the landscape.
Future Developments and Industry Shifts to Watch
Industry analysts expect increased scrutiny of the cartel’s dominance, potential regulatory interventions, and technological shifts that could decentralize compute access. Companies may seek alternatives to Nvidia’s chips or develop proprietary hardware, challenging the current power structure. Monitoring supply chain developments and contractual changes will be crucial to understanding how resilient this system remains.
Key Questions
Why do AI companies prefer renting compute rather than owning hardware?
Due to a GPU shortage and the high costs of building data centers, renting provides a faster and more flexible way to scale AI development without long lead times or massive capital expenditure.
How does Nvidia control the AI compute industry?
Nvidia supplies the majority of GPUs used in AI training, holds stakes in key firms, and controls chip allocation, giving it significant influence over who can access the hardware needed for AI development.
What risks does the current ‘neocloud’ cartel face?
The system’s reliance on circular financing and contractual dependencies makes it vulnerable to supply disruptions, legal or regulatory actions, and shifts in industry alliances.
Could this system lead to increased innovation or bottlenecks?
While centralized control may streamline certain aspects of AI development, it could also create bottlenecks and reduce competition, potentially slowing innovation if access becomes too restricted.
Source: ThorstenMeyerAI.com