📊 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 is increasingly renting compute from a small, interconnected group of firms, led by Nvidia. This creates a powerful but fragile cartel that controls access to critical hardware and funding.
Major AI companies are now leasing GPU compute from each other and a small group of landlords, with Nvidia at the center, creating a tightly linked cartel that controls access to critical AI infrastructure. This shift from ownership to leasing highlights a new power dynamic in the AI industry, with implications for market competition and supply chain stability.
According to recent analysis, the AI industry’s compute layer has evolved into a ‘neocloud’ — a hyperscale GPU rental market without legacy cloud baggage. Companies like CoreWeave, Meta, OpenAI, and xAI are leasing vast GPU capacity from Nvidia and each other, often on multi-billion dollar contracts.
In May 2026, xAI leased its supercomputer to Anthropic and Google, paying over $26 billion annually, despite having low utilization. This indicates that ownership of hardware has decoupled from AI development, with compute now primarily rented.
The financial flows reveal a circular pattern: Nvidia invests heavily in firms like OpenAI and Anthropic, financing their hardware needs while also holding equity stakes and pre-purchasing capacity. This interconnected financial web consolidates control over access to GPU resources, making a small number of firms the gatekeepers of AI compute.
Jensen Huang, Nvidia’s CEO, estimates the cost of a data center gigawatt at around $50 billion, with Nvidia capturing most of that revenue. The allocation of GPUs, therefore, becomes a critical lever of power, influencing which firms can scale AI models.
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.
The Power and Fragility of the AI Compute Cartel
This development means that a small group of firms, led by Nvidia, now controls the core hardware infrastructure for AI. Their ability to allocate compute resources effectively determines which companies can compete at scale. However, this concentration also introduces systemic risk: the entire AI development ecosystem becomes vulnerable if this cartel faces disruptions or fractures.
Furthermore, the circular financing and leasing model blurs lines of ownership and control, potentially limiting transparency and competition. The reliance on a handful of chip makers and landlords could influence AI innovation, pricing, and access for years to come.
Nvidia GPU cloud computing
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Origins of the AI Compute Rental Ecosystem
The trend toward renting compute began during the 2024–25 GPU shortage, which made ownership prohibitively expensive for many firms. As a result, companies turned to hyperscale GPU-as-a-service providers like CoreWeave, Meta, and others, creating a nascent market for GPU rental.
By 2026, this market had evolved into a tightly knit cartel, with Nvidia emerging as the dominant player. Notably, xAI’s leasing of its supercomputer to competitors marked a significant shift: even a self-described full-stack AI lab became a landlord, signaling a fundamental change in how compute resources are managed and controlled.
Prior to this, hardware was primarily owned outright or leased through traditional channels. The current model reflects a strategic move by Nvidia and others to centralize control over supply and pricing, effectively creating a choke point in the AI industry.
“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 leasing services
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unclear Risks and Future Stability of the Cartel
It is not yet clear how sustainable this tightly linked network of leasing and financing will be. The reliance on a small number of firms creates systemic vulnerabilities, but the full extent of potential disruptions or regulatory interventions remains uncertain. Additionally, the long-term implications for market competition and innovation are still developing.
high performance GPU servers
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Potential Breakpoints and Industry Shifts Ahead
Next steps include monitoring how regulatory bodies respond to this concentration of power, especially if supply disruptions or anti-trust concerns arise. Further, the industry may see efforts to diversify supply chains or develop alternative hardware solutions. The evolution of lease agreements and ownership models will also influence the future landscape of AI compute infrastructure.
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
Why is Nvidia so central to the AI compute industry?
Nvidia dominates GPU supply and controls a significant share of the market, investing heavily in AI firms, and controlling hardware allocation through its contracts and equity stakes.
What does it mean for AI companies to rent compute instead of owning it?
It means that access to critical hardware is now governed by leasing agreements, which can be revoked or re-priced, shifting power from owners to landlords and financiers.
Could this cartel structure limit competition or innovation?
Yes, the concentration of control over hardware access could restrict smaller firms’ ability to scale AI models, potentially impacting market competition and technological progress.
What risks does this interconnected leasing network pose?
The system’s fragility could lead to supply shocks if key firms face disruptions, and the high level of financial interdependence increases systemic risk.
How might regulators respond to this development?
Regulators could investigate anti-trust concerns or impose rules to limit concentration, but such actions depend on evolving industry dynamics and policy priorities.
Source: ThorstenMeyerAI.com
