📊 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 major firms rent compute from a tightly linked group of GPU landlords, forming a cartel controlled mainly by Nvidia. This setup influences market power, funding, and supply chains, but also introduces fragility.
Major AI companies now rely on a small, interconnected group of GPU landlords to rent their compute capacity, with Nvidia emerging as the dominant player. This shift signifies a fundamental change in how AI infrastructure is owned and controlled, impacting market dynamics and supply chains.
Recent reports reveal that the AI industry’s compute layer functions less like a free market and more like a tightly controlled cartel. Companies such as OpenAI, Anthropic, Meta, and xAI are leasing vast amounts of GPU capacity from a handful of firms, with Nvidia at the core. For instance, xAI leased its Colossus 1 supercomputer to Anthropic for about $1.25 billion per month and to Google for approximately $920 million per month. This pattern illustrates a shift where ownership of hardware has decoupled from its use, with leasing becoming the dominant model.
Furthermore, the financial flows reveal a circular pattern: Nvidia has invested heavily in AI firms, including a $100 billion commitment to OpenAI, and holds equity stakes in multiple companies. Major deals involve hundreds of billions of dollars, such as OpenAI’s projected $1.15 trillion expenditure on compute over a decade, spread across a small group of suppliers like Broadcom, Oracle, Microsoft, and Nvidia. These arrangements create a tightly linked network where a few firms control the flow of chips, funding, and capacity, effectively forming a choke point in the AI ecosystem.
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 the AI Compute Cartel
This concentration of compute resources in a small circle of firms gives Nvidia and its partners significant market power, influencing AI development, pricing, and access. It also creates a dependency that could lead to vulnerabilities if supply chains or financial arrangements are disrupted. The model’s circular financing and leasing structure make the entire system fragile, risking sudden shifts if key players withdraw or if regulatory pressures increase.
HPE NVIDIA Tesla V100 32GB HBM2 PCIe 3.0 x16 Passive GPU Computational Accelerator for AI Machine Learning HPC Deep Learning 699-2G500-0216-400 (Renewed)
NVIDIA Volta GV100 Architecture — 5,120 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Origins and Evolution of AI Compute Concentration
The trend toward renting compute emerged around 2024–25, driven by a GPU shortage that made owning hardware less feasible for most AI labs. Companies like CoreWeave, Meta, and OpenAI began leasing large GPU clusters, with the market evolving into a de facto cartel by 2026. Notably, xAI’s leasing of its supercomputer to rivals marked a turning point, as it signaled a move toward self-sourcing and leasing capacity to others, blurring the lines between owners and renters.
Prior to this, the AI industry relied on traditional cloud providers and hardware ownership. The current model, characterized by leasing and circular financing, was prompted by supply shortages and the high costs of building dedicated infrastructure. Nvidia’s strategic investments and control over chip allocation have cemented its central role, transforming the compute layer into a chokepoint with global implications.
“A gigawatt of AI data center capacity costs about $50 billion, and most of that flows directly to Nvidia.”
— Jensen Huang, Nvidia CEO
Graphics Processing Unit-Based High Performance Computing in Radiation Therapy (Series in Medical Physics and Biomedical Engineering)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Risks Does the AI Compute Cartel Face?
It remains unclear how stable this tightly linked network of leasing and financing will prove long-term. Potential disruptions include regulatory actions, supply chain shocks, or shifts in company strategies that could weaken Nvidia’s control or fragment the cartel. The fragility of the circular financing model poses risks that are not yet fully understood or tested under stress.
Waiting for Free Compute Financial Tech Hardware Humor AI T-Shirt
Features an empty wallet with a moth flying out. A humorous take on the high cost of high-performance…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Developments in AI Hardware Control
Expect increased scrutiny from regulators as the concentration of compute capacity raises antitrust concerns. Nvidia’s role as both supplier and investor could face challenges, potentially leading to calls for more open or diversified hardware markets. Additionally, as AI models grow larger and more complex, the demand for compute will intensify, possibly prompting new alliances or shifts in leasing arrangements.
Monitoring regulatory responses and supply chain resilience will be crucial in understanding how this cartel evolves or dissolves in the coming years.
NComputing RX440(RDP) Thin Client Built on Raspberry Pi4 for Microsoft AVD, Windows 365 Cloud PC, RDS, Verde VDI, vSpace Pro Enterprise
Requires connection license for specific virtualization platform you intent to use (Not Included)
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 market?
Nvidia dominates because it supplies the majority of high-performance GPUs used in AI training and inference, controls chip allocation, and has invested heavily in AI firms, creating a network of financial and hardware dependencies.
How does leasing compute from a few firms affect AI development?
It concentrates power among a small group of suppliers, which can influence pricing, access, and innovation, potentially limiting competition and creating vulnerabilities if the supply chain is disrupted.
Could regulatory action break up this AI compute cartel?
It is possible, especially if authorities view the concentration of control as anti-competitive. However, the current structure’s complexity and the strategic investments involved make regulation challenging.
What risks does this cartel pose to the AI industry?
The main risks include supply chain disruptions, increased costs, and reduced competition, which could slow AI innovation or cause sudden market shifts if key players withdraw or face regulatory penalties.
Will ownership of hardware return to a more distributed model?
It is uncertain. Currently, leasing dominates due to supply shortages and high costs, but technological advancements or regulatory pressures could shift the balance back toward ownership in the future.
Source: ThorstenMeyerAI.com
