📊 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
The AI industry has shifted to a model where companies rent compute from each other, creating a small, interconnected cartel. Nvidia dominates this system, holding critical control over GPU supply and financing.
In 2026, the AI industry has transitioned to a model where companies rent their compute infrastructure from each other, rather than owning it outright. This shift, driven by a GPU shortage and high demand for AI training, has resulted in a network of companies with significant interdependencies, with Nvidia playing a central role in the supply and distribution of GPU resources. This shift, driven by a GPU shortage and high demand for AI training, has resulted in a network of companies with significant interdependencies, with Nvidia playing a central role in the supply and distribution of GPU resources.
Major AI firms such as OpenAI, Anthropic, and xAI are leasing hundreds of millions of dollars worth of GPU capacity from each other, often on multi-billion dollar contracts. This interconnected system is partly explained by the industry’s evolving structure discussed in The Neocloud Cartel. Notably, xAI leased its supercomputer to competitors like Anthropic and Google, signaling a move toward self-reliance in infrastructure while still participating in the rental ecosystem. This pattern underscores the industry’s reliance on a small circle of GPU landlords, primarily Nvidia, which supplies the majority of the chips used in AI training.
Financial flows reveal a circular system: Nvidia, for example, invested up to $100 billion in OpenAI’s buildout and holds equity in multiple firms involved in AI infrastructure. Companies like Microsoft, Amazon, and AMD also participate heavily, with multi-billion dollar commitments, often financed by the chipmaker itself. The contracts are replete with clauses that allow Nvidia to control GPU allocation, effectively granting it the power to gate access and influence market dynamics.
This network of leasing, financing, and equity stakes has formed a de facto arrangement—an oligopoly where a small number of firms have significant influence over AI compute access. For more on this industry dynamic, see The Neocloud Cartel. The system’s design makes access to hardware subject to contractual terms, which can impact availability and pricing, raising questions about the industry’s long-term stability and competition.
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 a Compute Cartel for AI Development
The emergence of this compute arrangement concentrates decision-making power among a limited number of firms, which could influence competition and innovation. The reliance on leasing and interconnected financing introduces dependencies that could be affected by supply chain issues or contractual disputes. This structure also raises questions about market transparency and the potential for market dominance, which could influence AI development trajectories and pricing.
NVIDIA Jetson Orin Nano Super Developer Kit
The NVIDIA Jetson Orin Nano Developer Kit sets a new standard for creating entry-level AI-powered robots, smart drones,…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Rapid Growth and Centralization in AI Compute Infrastructure
Over the past three years, the AI industry has shifted from owning its hardware to relying on leasing arrangements, driven by a GPU shortage and the costs associated with building data centers. Companies like CoreWeave, Meta, and OpenAI have relied on Nvidia’s hardware, with Nvidia’s investments and strategic contracts reinforcing its market position. This trend reflects a broader pattern of industry centralization, where a few firms control critical infrastructure and financing, creating a concentrated environment for AI compute resources.
Historically, hardware ownership was more dispersed across many firms; now, the infrastructure is concentrated within a small, interconnected network. Leasing agreements often include clauses that provide landlords with certain rights over capacity, which can influence availability and control, further consolidating Nvidia’s role in the ecosystem.
“A gigawatt of AI data center capacity costs roughly $50 billion, and Nvidia captures a significant portion of those dollars.”
— Jensen Huang, Nvidia CEO
high performance AI compute server
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unclear Risks and Potential Disruptions to the Cartel
The stability of this arrangement depends on various factors, including supply chain conditions, financial health of key players, and regulatory developments. Disruptions or shifts in contractual terms could influence Nvidia’s control and the overall ecosystem, but the long-term resilience of this structure remains uncertain.
GPU leasing hardware
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Developments and Regulatory Scrutiny of AI Compute Monopoly
Regulatory authorities may increase oversight of Nvidia and related leasing practices, particularly if concerns about market dominance arise. Technological innovations or alternative compute architectures could also influence the current system. Companies might seek to diversify their supply chains or develop proprietary infrastructure to reduce reliance on existing arrangements.
In the near term, ongoing contractual agreements and strategic investments are likely to sustain the current structure, while uncertainties about future stability persist.
Deep Learning at Scale: At the Intersection of Hardware, Software, and Data
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How does Nvidia control access to AI compute resources?
Nvidia influences access primarily through its dominant supply of GPUs, contractual provisions, and allocation decisions that can prioritize or restrict capacity for certain clients, effectively serving as a key gatekeeper in the AI compute ecosystem.
Why are companies leasing compute instead of owning hardware?
The high costs of building and maintaining data centers, along with shortages of GPUs, have made leasing a practical approach for many AI firms to scale their operations efficiently.
What risks does this compute arrangement pose to the AI industry?
Dependence on a limited number of suppliers and contractual control mechanisms could lead to supply constraints, reduced competition, and higher prices, which may impact the pace and cost of AI development.
Could regulatory action influence this arrangement?
Regulatory scrutiny could increase, especially if concerns about market concentration and competitive fairness arise. The outcome will depend on legal and policy developments in the coming years.
Source: ThorstenMeyerAI.com
