📊 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 GPU compute from each other, creating a tightly interconnected cartel led by Nvidia. This system concentrates power and control, but also introduces vulnerabilities.
In May 2026, xAI leased its Colossus 1 supercomputer to Anthropic for approximately $1.25 billion per month and to Google for about $920 million per month. This marks a pivotal moment in the AI industry, illustrating how firms increasingly rent compute from each other rather than owning it outright, effectively creating a small, interconnected cartel centered around Nvidia.
Major AI companies, including OpenAI, Meta, and xAI, now rely heavily on rented GPU compute, with contracts often exceeding billions of dollars annually. This shift was driven by a GPU shortage in 2024–25, which made owning hardware impractical, leading firms to rent from a small group of GPU landlords such as CoreWeave, Nvidia, and others.
The core of this system is a tightly interconnected network of firms financing each other’s compute needs through circular investments and leasing agreements. Nvidia, as the dominant supplier, controls the majority of GPU capacity and financial flows, effectively holding the choke point in the industry’s compute infrastructure. Nvidia’s investments in firms like OpenAI and its control over chip allocation give it outsized influence over AI development and deployment.
In this setup, firms like xAI have become both users and landlords, leasing their hardware to others, further entrenching the cycle of dependency and control. The contracts often include governance clauses, such as xAI’s lease to Anthropic, which reserves the right to reclaim capacity if certain conditions are met, turning supply agreements into control mechanisms.
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 Concentrated Compute Cartel
This emerging system consolidates power within a small circle of firms, primarily Nvidia, which controls access to AI compute resources. The model creates a fragile but dominant choke point, where Nvidia’s decisions on chip allocation and financing influence the entire AI ecosystem. While this concentration enables rapid scaling and investment, it also poses risks of supply shocks and increased dependency, potentially impacting innovation, competition, and pricing in AI development.
GPU server rental
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Rapid Growth and Centralization of AI Compute Resources
Between 2024 and 2026, the AI industry faced a significant GPU shortage, prompting a shift from ownership to renting compute. Major players like OpenAI, Meta, and xAI have committed hundreds of billions of dollars to GPU leasing, primarily from Nvidia and other hyperscalers like CoreWeave. This period saw the rise of the ‘neocloud’ model—AI-specific hyperscalers offering GPU-as-a-service without traditional cloud baggage—further accelerating the centralization of compute resources.
The formation of this cartel-like network was facilitated by circular financing arrangements and strategic investments, with Nvidia emerging as the central hub. Nvidia’s investments in firms like OpenAI and its control over chip supply have turned it into the gatekeeper of AI compute capacity, effectively making it the choke point in the industry’s infrastructure.
“The cost of a gigawatt of AI data center capacity is roughly $50 billion, with about $35 billion flowing to Nvidia.”
— Jensen Huang, Nvidia CEO
INFINIBAND FOR HIGH-PERFORMANCE COMPUTING AND AI CLUSTERS: Configure RDMA networking, optimize GPU interconnects, and build low-latency infrastructure for distributed training and HPC workload
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unclear Risks and Potential Fragilities of the Cartel
While the system consolidates power and accelerates AI development, it also introduces vulnerabilities. The reliance on a small number of suppliers and the circular financing structure could lead to supply shocks if any link in the chain faces disruption. It remains unclear how resilient this cartel is to regulatory interventions, market shocks, or technological shifts that could alter the balance of power.
NVIDIA and the AI Revolution: How GPUs Became the Most Important Technology on Earth
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Developments and Potential Industry Shifts
Industry observers expect increased scrutiny of Nvidia’s dominance and the cartel-like structure of AI compute markets. Regulatory agencies might investigate anti-competitive practices, and alternative hardware or decentralized compute solutions could challenge the current model. Additionally, firms may seek to diversify their compute sources to reduce dependency on a single supplier, potentially fracturing the current concentration of power.
The DOJO Supercomputer: Elon Musk’s High-Stakes AI Revolution: The Untold Story of Tesla’s Bold Gamble to Dominate the Future of Artificial … TECH, … TECH, SCIENECE AND SPACE TREND UPDATES)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why do AI companies prefer renting compute instead of owning hardware?
Due to the GPU shortage in 2024–25, renting became the only practical way to scale AI training and inference without long delays. It also allows flexibility and reduces upfront capital expenditure.
What role does Nvidia play in this AI compute cartel?
Nvidia is the dominant supplier, controlling most GPU capacity and financing. It influences pricing, allocation, and investment flows, effectively holding the choke point in the industry.
Could this concentration of control harm AI innovation?
Yes, the reliance on a small group of firms and the potential for supply disruptions could limit competition and slow innovation, while also increasing prices and dependency risks.
Are there alternatives to this rental-based compute model?
Potential alternatives include decentralized compute networks, custom hardware development, or increased investment in owned infrastructure, but these face significant technical and economic barriers.
Source: ThorstenMeyerAI.com
