NVIDIA is starting to take a cut from cloud providers' revenue
- Core Thesis: NVIDIA is evolving from a chip vendor into the "central bank" of the AI computing ecosystem. By providing financial guarantees for emerging cloud service providers through a GPU capacity leaseback model, it secures a share of their revenue, thereby lowering customer financing barriers and expanding market control.
- Key Elements:
- NVIDIA launched the "AI Computing Cooperation Plan," promising to lease back unsold GPU capacity from emerging cloud service providers at agreed-upon prices, serving as a credit endorsement.
- In exchange, NVIDIA receives a gradually decreasing percentage share of these cloud providers' revenue, creating recurring income linked to usage volume.
- The initial participants include cloud service provider Sharon AI (planning to deploy 40,000 GB300 GPUs) and Firmus (building a 360-megawatt AI factory in Indonesia).
- This model aims to solve the core challenge where emerging AI companies struggle to secure GPU financing due to low credit ratings, accelerating their infrastructure buildout.
- This move is a continuation of NVIDIA's strategy to reduce dependence on major clients like Amazon and Microsoft, which are developing their own competing AI chips.
- NVIDIA has already invested billions of dollars in companies like CoreWeave through equity investments and capacity leasebacks, having previously committed $6.3 billion to purchase unsold capacity from CoreWeave.
- NVIDIA injected an additional $3.5 billion to guarantee client data center leases in exchange for stock purchase rights, establishing a multi-layered interest alignment mechanism.
Original Author: Dong Jing
Original Source: Wall Street CN
Nvidia is leveraging its powerful balance sheet as a market lever, providing financial backing to emerging cloud service providers in exchange for revenue sharing, quietly evolving from a chip seller into the "central bank" of the AI computing ecosystem.
On July 1st, according to a report by tech media outlet The Information, Nvidia is offering financial backstop commitments to young cloud service providers that lease its GPUs — if these companies cannot find enough AI developers to rent their computing power, Nvidia will repurchase their unsold GPU capacity at an agreed-upon price.
In exchange, Nvidia will take a percentage cut of these cloud service providers' revenue, with the share gradually decreasing over the contract term. GPU cloud service providers Firmus and Sharon AI have already joined this program, and three other executives with business relationships with Nvidia have confirmed the arrangements.
On July 1st, Nvidia announced on its official website the launch of a new business model combining revenue sharing with credit support. This model allows AI cloud providers to procure Nvidia infrastructure without fully bearing the upfront capital expenditure themselves, and to offer computing services downstream to AI-native enterprises, model developers, and corporate clients.
The report states that some within Nvidia have internally called this program the "AI Compute Partnership." An Nvidia spokesperson also confirmed the program's existence. This move marks a significant strategic shift for Nvidia:
On one hand, it expands the customer base by lowering the financing barrier for emerging cloud service providers. On the other hand, through revenue sharing, it directly participates in the profit distribution of the downstream computing market, extending its control further down the AI industry chain.
Model Shift: From Selling Chips to Sharing Cloud Revenue
According to Nvidia's official press release, beyond standard product revenue, Nvidia will additionally share in the cloud service revenue, thereby creating a recurring income stream linked to usage volume. The core intention of this model is to break down the financing barriers that have historically constrained startups from obtaining large-scale AI computing power.
Nvidia positions this framework as the "DSX AI Factory" model, targeting AI service scenarios requiring continuous cross-regional operation, high utilization rates, and multi-tenant accelerated computing.
Sharon AI and Firmus are the first cloud providers to participate in this model. Sharon AI plans to deploy up to 40,000 Nvidia Grace Blackwell GB300 GPUs. Firmus is building a DSX AI Factory campus in Batam, Indonesia, expected to expand to 360 megawatts and house up to 170,000 Nvidia GPUs. These two deployments directly reflect Nvidia's latest progress in translating computing demand into fundable, actionable infrastructure.
Nvidia points out that emerging AI companies have historically faced severe limitations in accessing capital-intensive infrastructure — even signing long-term commitment contracts often isn't enough to secure financing for hardware procurement. This means numerous AI-native companies, model developers, and inference service providers face lengthy waits when scaling their computing capabilities: site selection, power procurement, construction, and hardware debugging — each step can take months or even longer.
The promise of the new model is this: by realigning the economic structure, these groups can gain access to full-stack accelerated computing capabilities much faster, without waiting for the completion of traditional infrastructure construction cycles.
The Logic of the Backstop: Solving the Core Problem of GPU Financing
According to reports, GPUs are typically the most expensive component in AI data centers. For chip buyers with lower credit ratings, obtaining sufficient loans is a major hurdle in itself.
A data center executive commented on this, saying Nvidia's type of transaction "kills two birds with one stone." He explained that if Nvidia only guarantees a lease for the data center facility, "you still have the problem of 'how to finance the GPUs.'" But if Nvidia commits to paying for the facility's unsold computing capacity, "the GPU financing problem is solved, and so is the data center's financing problem."
In other words, Nvidia's backstop commitment essentially acts as a credit enhancement tool, enabling emerging cloud service providers, who would otherwise struggle to secure bank loans, to leverage larger amounts of capital and accelerate data center construction.
Strategic Intent: Breaking the Monopoly of Major Customers
Nvidia is launching this series of initiatives against a clear strategic backdrop. Currently, a small group of large cloud service providers like Amazon, Microsoft, SpaceX, Oracle, Meta, and Google purchase most of Nvidia's chip production. However, several of these companies are developing competing AI chips in-house, posing a potential threat to Nvidia.
To reduce its dependence on these giant customers, Nvidia has been cultivating emerging GPU cloud service providers like CoreWeave for several years. This "AI Compute Partnership" represents a continuation and deepening of that strategy.
According to a previous report by The Information, Nvidia recently has also been in talks to provide a financial guarantee for OpenAI to lease a large data center in Ohio. The total cost of building out this data center at current chip, labor, power, and other material prices could reach up to $500 billion.
Capital Influx: From Equity Investments to Capacity Guarantees
Nvidia's financial commitment in this direction is already substantial.
To date, Nvidia has invested billions of dollars in several emerging cloud service providers in exchange for equity and, in some cases, agreed to lease back chips from these companies. This involves firms like CoreWeave and Lambda, with the total transaction value reaching tens of billions of dollars. According to a previous report by The Information, Nvidia's own researchers have used GPU servers leased back from Lambda.
In terms of capacity guarantees, Nvidia began pushing forward with related transactions last fall. In September 2024, Nvidia committed to buying all of CoreWeave's unsold capacity until 2032 if CoreWeave could not find tenants, at a contract value of $6.3 billion. This move effectively alleviated investor concerns about CoreWeave's high-leverage business model, boosting its share price by nearly 30% in the following week.
According to a regulatory filing by Nvidia in May (covering the quarter through April), Nvidia subsequently added another $3.5 billion to guarantee customer data center leases in exchange for the right to purchase their stock.
In summary, Nvidia is building a multi-layered interest-binding mechanism: equity investments, capacity leasebacks, lease guarantees, and now, revenue sharing. Each layer deepens the financial ties between Nvidia and its downstream cloud service providers, allowing Nvidia to directly share in the incremental revenue from AI computing commercialization, beyond just chip sales.
