📊 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
In 2026, AI firms are increasingly renting compute from each other, forming a cartel centered on Nvidia. This shift decouples ownership from use, raising questions about market power and stability.
In 2026, the AI industry has shifted to a model where most companies rent their compute resources from a small group of landlords, primarily Nvidia, rather than owning their own hardware. This development signifies a major change in how AI infrastructure is accessed and controlled, with profound implications for market power and competition.
Recent reports indicate that the AI sector is increasingly relying on a network of GPU landlords, known as the ‘neocloud,’ which includes companies like CoreWeave, Meta, OpenAI, and xAI. Notably, xAI leased its supercomputer to competitors Anthropic and Google for over $26 billion annually, despite owning minimal utilization of its hardware. This shift reflects a decoupling of hardware ownership from AI development, driven by GPU shortages and the high costs of building proprietary infrastructure.
Furthermore, a circular flow of money and hardware has emerged, with major players such as Nvidia, Microsoft, and AMD investing billions into AI firms like OpenAI and Anthropic. Nvidia, in particular, acts as the central choke point, controlling the majority of GPU supply and holding equity stakes in key companies. Nvidia’s investments and allocation decisions effectively determine who can access AI compute resources, giving it outsized influence over the industry. This interconnected financing and hardware leasing create a small cartel where access is repriceable, revocable, and dependent on contractual and supply chain power.
While this structure offers flexibility and rapid scaling, it also introduces fragility. The circular dependency means that disruptions in Nvidia’s supply chain or policy decisions could cascade through the industry, threatening stability. The dependence on a few large firms for capital and hardware further concentrates power, raising concerns about monopolistic tendencies and market resilience.
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 Small AI Compute Cartel
This development matters because it concentrates control over AI infrastructure in the hands of a few dominant firms, primarily Nvidia. The decoupling of ownership from use enables rapid scaling but also creates systemic risks. If Nvidia or other key players face disruptions, the entire AI industry could be affected, highlighting vulnerabilities in the current model. The formation of this cartel could influence pricing, access, and innovation in AI for years to come, raising questions about competition and market fairness.
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.
Rise of the Neocloud and Market Concentration
Over the past three years, the AI hardware market has undergone a transformation driven by GPU shortages and the high costs of building proprietary infrastructure. Companies like CoreWeave, Meta, and OpenAI have increasingly relied on renting GPU capacity from Nvidia and other landlords, with contracts often exceeding billions of dollars annually. The emergence of the ‘neocloud’—a dedicated hyperscaler for AI—has facilitated this shift, enabling rapid scaling without ownership. In May 2026, xAI’s leasing of its supercomputer to rivals marked a turning point, illustrating how hardware leasing has become a core strategic element rather than a secondary consideration.
This interconnected web of financing and leasing has led to a small group of firms controlling the flow of compute, with Nvidia at the center due to its dominant supply and investment role. The circular flow of capital and hardware has created a de facto cartel, where access to compute is governed by contracts, supply allocations, and financial dependencies, rather than open market competition.
“The cost of a gigawatt of AI data center capacity is roughly $50 billion, with the majority flowing directly to Nvidia.”
— Jensen Huang, Nvidia CEO
AI hardware leasing platforms
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unclear Risks and Potential Disruptions in the Cartel
It remains unclear how vulnerable this tightly interconnected system is to disruptions, such as supply chain shocks, regulatory actions, or shifts in corporate strategy. The fragility of the circular financing and leasing model could pose systemic risks, but specific scenarios and their likelihood are still being evaluated.
enterprise GPU rental services
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Developments and Industry Responses
Industry analysts expect ongoing consolidation and possible regulatory scrutiny as the cartel’s influence grows. Nvidia’s role as a gatekeeper may come under pressure, especially if new hardware technologies or alternative supply chains emerge. Companies may also seek to diversify their hardware sources or develop proprietary infrastructure to reduce dependence on this small group of landlords. Monitoring Nvidia’s supply policies and investment strategies will be key to understanding how the market evolves in the coming months.
AI Data Center Infrastructure Engineering: Power Distribution, Liquid Cooling, High-Density Networking, and Energy Efficiency for GPU Training … Hardware & Compiler Engineering Series)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How does Nvidia control the AI compute market?
Nvidia controls the market primarily through its dominant GPU supply, strategic investments, and allocation decisions. It acts as the central gatekeeper, determining which firms get access to hardware and at what cost.
Why are companies renting compute instead of owning hardware?
GPU shortages, high capital costs, and the need for rapid scaling have made renting the most practical option for AI firms, allowing flexibility without long-term ownership commitments.
What risks does this cartel pose to the AI industry?
The primary risks include supply chain disruptions, market monopolization, and systemic fragility if Nvidia or other key players face operational or regulatory challenges.
Could this structure lead to anti-competitive behavior?
Yes, the concentration of control over hardware access and financing could raise antitrust concerns, especially if it stifles competition or innovation.
What might change in the industry if Nvidia’s role diminishes?
If Nvidia’s influence wanes or new supply sources emerge, the industry could see increased competition, more diverse hardware options, and potentially a more resilient market structure.
Source: ThorstenMeyerAI.com
