Nvidia has paused some revenue-sharing deals under a financing initiative designed to help smaller artificial-intelligence cloud companies purchase its expensive AI chips, according to a Wall Street Journal report cited by Reuters. The chipmaker stepped back from the program last week, less than two months after introducing it, amid concerns about potential antitrust scrutiny and the extent of control Nvidia could exert over customers’ businesses. The company said the broader business model remains in place and is evolving because of strong demand.
The initiative was designed to go beyond Nvidia’s traditional role as a chip supplier. Under the proposed arrangement, Nvidia could provide credit support to AI cloud companies, earn revenue from selling them GPUs and then receive a share of the cloud revenue generated from Nvidia-powered computing capacity. Nvidia also sought to act as a potential buyer of unused compute capacity, helping cloud providers secure financing while reducing lenders’ concerns about future utilization. The model had the potential to generate billions of dollars in additional revenue for Nvidia, but it also raised questions about circular financing, competition and the chipmaker’s growing influence over the AI infrastructure ecosystem.
Nvidia Pauses AI Cloud Revenue-Sharing Deals
Nvidia has paused some agreements that were being negotiated under its new financing initiative for AI cloud companies.
The program was introduced in July to help smaller AI infrastructure providers overcome one of their biggest obstacles: the enormous upfront cost of purchasing Nvidia’s latest-generation GPUs.
Instead of simply selling chips, Nvidia sought to help cloud companies finance those purchases and then participate in the revenue generated when the GPUs were rented to AI customers.
The reported pause does not mean Nvidia has abandoned the broader strategy.
An Nvidia spokesperson said the business model introduced in July remains in place and is continuing to evolve in response to strong demand.
Nvidia Revenue-Sharing Initiative At A Glance
| Particular | Details |
|---|---|
| Company | Nvidia |
| Initiative | AI cloud financing / revenue sharing |
| Launch | July 2026 |
| Latest development | Some deals paused |
| Timing of pause | Last week |
| Target customers | Smaller AI cloud companies |
| Nvidia role | Chip supplier + financing support |
| Revenue model | Hardware sales + share of cloud revenue |
| Proposed revenue share | 50% above a specified threshold |
| Potential issue | Antitrust scrutiny |
| Current status | Could be revised or folded into another program |
| Nvidia’s position | Broader model remains in place |
The distinction between pausing some deals and abandoning the overall strategy is important. Nvidia is still seeking ways to expand access to its chips through alternative financing and infrastructure arrangements.
How Nvidia’s Revenue-Sharing Model Works
The proposed structure was designed to address the financing challenge faced by AI cloud companies.
These companies need to purchase expensive GPUs before they can rent computing capacity to customers. That creates a large capital requirement and exposes them to the risk that demand may not be strong enough to keep the hardware fully utilized.
Nvidia’s model attempted to address both problems.
Nvidia Sells AI Chips
↓
AI Cloud Company Purchases GPUs
↓
Nvidia Provides Credit / Financing Support
↓
Cloud Company Builds AI Capacity
↓
Customers Rent Compute
↓
Cloud Company Generates Revenue
↓
Nvidia Receives A Share Of Revenue
Nvidia also sought the right to rent back unused capacity in certain arrangements, effectively providing a potential buyer for excess computing capacity.
That could make it easier for the cloud company to secure debt financing because lenders would have greater confidence that some of the infrastructure could generate revenue even if the company’s own customer demand was insufficient.
Nvidia Could Earn Twice From The Same AI Infrastructure
The model was unusual because it potentially gave Nvidia two revenue streams from the same hardware.
First, Nvidia would receive money when it sold GPUs to the cloud provider.
Second, it could receive a portion of the cloud revenue generated by those Nvidia-powered GPUs.
Traditional Vs Revenue-Sharing Model
| Model | Nvidia Revenue |
|---|---|
| Traditional chip sale | GPU sale |
| Revenue-sharing model | GPU sale + share of cloud revenue |
| Nvidia capacity backstop | Potential additional utilization |
| Strategic benefit | Greater ecosystem involvement |
The Journal reported that Nvidia would receive 50% of cloud-provider revenue above a specified threshold under proposed deals.
The arrangement could therefore turn Nvidia from a hardware supplier into a participant in the economics of the AI cloud businesses using its chips.
Why Nvidia Paused The Deals
One of the biggest concerns is antitrust scrutiny.
Some Nvidia employees reportedly warned current and prospective customers that the structure could attract regulatory attention.
The issue is Nvidia’s increasing influence over both the supply of AI chips and the businesses that depend on those chips.
If Nvidia can determine which customers receive GPU capacity, dictate how cloud providers rent their hardware and receive a share of their revenues, regulators could potentially question whether the arrangement restricts competition.
Potential Regulatory Concerns
| Concern | Why It Matters |
|---|---|
| Market power | Nvidia dominates AI accelerators |
| Customer restrictions | Nvidia could influence who gets access to GPUs |
| Revenue sharing | Supplier participates in customer economics |
| Capacity allocation | Nvidia could influence cloud-provider strategy |
| Approved customers | Could restrict downstream competition |
| Vertical integration | Hardware supplier moves into cloud economics |
| Circular financing | Raises questions about artificial demand |
The reported pause suggests Nvidia is taking those concerns seriously as it expands its role across the AI infrastructure stack.
Nvidia Was Seeking Greater Control Over Cloud Capacity
The Journal reported that Nvidia told some cloud providers they could rent its chips only to customers approved by Nvidia.
The company also preferred that cloud capacity be distributed across multiple smaller AI firms rather than concentrated in one large customer.
Such provisions could help Nvidia expand the number of companies using its hardware, but they also create potential questions about how much control a chip supplier should have over a customer’s commercial decisions.
Proposed Nvidia Controls
Nvidia GPUs
↓
Cloud Provider
↓
Nvidia Approval
↓
Eligible Customers
↓
Compute Rentals
↓
Revenue Sharing
The more extensive these controls become, the more the relationship starts to resemble a strategic operating partnership rather than a conventional hardware transaction.
Smaller AI Cloud Companies Are The Main Target
The initiative was aimed primarily at smaller AI cloud providers.
These companies are attempting to compete with Amazon Web Services, Microsoft Azure and Google Cloud by offering dedicated AI computing capacity, often using Nvidia GPUs.
Their advantage can be specialization and faster deployment, but their disadvantage is the enormous amount of capital required to purchase and operate AI infrastructure.
AI Cloud Provider Model
| Requirement | Challenge |
|---|---|
| Nvidia GPUs | Very high upfront cost |
| Data centers | Significant capital expenditure |
| Electricity | Rising infrastructure cost |
| Networking | Additional hardware investment |
| Customer contracts | Needed to support financing |
| GPU utilization | Must remain high |
| Debt servicing | Depends on recurring revenue |
Nvidia’s financing model was designed to address several of these constraints simultaneously.
AI Cloud Companies Face Huge Capital Requirements
The AI infrastructure boom has created a new category of cloud companies that specialize in providing GPU compute to AI developers.
Companies such as CoreWeave, Lambda, Nebius and others have raised substantial capital to build GPU-heavy data centers.
CoreWeave, for example, has expanded rapidly and has secured billions of dollars of financing while working closely with Nvidia. It also received a $2 billion investment from Nvidia in early 2026 and has expanded its relationship with the chipmaker to build more than 5 gigawatts of AI factories by 2030.
This demonstrates why Nvidia’s financing strategy can be attractive to both sides.
AI Demand
↓
More GPUs Needed
↓
Cloud Provider Needs Capital
↓
Nvidia Helps Finance GPUs
↓
More Nvidia Chips Deployed
↓
More AI Compute Revenue
The concern is that this cycle can potentially reinforce Nvidia’s already dominant market position.
Nvidia’s AI Ecosystem Financing Has Become Much Larger
The paused revenue-sharing program is only one part of Nvidia’s broader effort to support the AI infrastructure ecosystem.
This month, Nvidia helped arrange roughly $500 billion in financing from major U.S. financial institutions for its customers, according to Reuters. It has also agreed to guarantee up to $105 billion to help OpenAI lease a large data center.
Nvidia’s Broader Financing Strategy
| Initiative | Reported Scale |
|---|---|
| Financing arranged for customers | ~$500 billion |
| OpenAI data-center guarantee | Up to $105 billion |
| CoreWeave investment | $2 billion |
| Proposed revenue-sharing model | Share of cloud revenue |
| AI cloud capacity support | Multiple providers |
These arrangements show that Nvidia is increasingly involved in financing the infrastructure needed to consume its own chips.
Circular Financing Concerns Are Growing
The financing model has attracted scrutiny because of the possibility of circular transactions.
A simplified circular structure could look like this:
Nvidia
↓
Provides Financing
↓
AI Cloud Company
↓
Buys Nvidia GPUs
↓
AI Customer
↓
Pays Cloud Provider
↓
Cloud Provider Shares Revenue With Nvidia
In some circumstances, Nvidia can therefore help finance the purchase of its own products while participating in the revenue generated from those products.
That does not necessarily mean the transactions are artificial or improper. However, the structure can raise questions about how much of the reported demand for AI infrastructure is ultimately supported by capital originating from Nvidia itself.
The Reuters report specifically noted investor concerns over Nvidia’s role in so-called circular deals that could artificially inflate demand.
Nvidia Says The Model Could Generate Billions
Despite the reported pause, Nvidia has not backed away from the economic opportunity.
During its latest earnings call, the company said the new model could generate billions of dollars in revenue over the medium to long term.
That potential is understandable given the scale of the AI infrastructure market.
Nvidia reported $96.22 billion in quarterly revenue for the latest quarter, while its data-center business generated approximately $89 billion.
Nvidia’s Latest Scale
| Metric | Latest Reported Figure |
|---|---|
| Quarterly revenue | $96.22 billion |
| Data-center revenue | ~$89 billion |
| Net income | $59.69 billion |
| Quarterly revenue growth | More than 2X YoY |
| Current-quarter revenue guidance | ~$108 billion |
| FY2028 revenue-growth outlook | ~70% |
The enormous size of Nvidia’s core business means the company can afford to experiment with new financing structures, but it also means regulators and investors have greater reason to examine those arrangements.
Nvidia’s Core Business Remains Extremely Strong
The pause in financing deals should not be interpreted as a weakening in demand for Nvidia chips.
The company reported record quarterly sales of $96.22 billion, exceeding Wall Street’s $92.27 billion consensus estimate.
Its data-center business generated $89 billion, more than double the year-earlier figure. Nvidia also forecast roughly $108 billion in revenue for the following quarter.
Nvidia shares subsequently jumped 8.7% on Thursday after the company provided an optimistic longer-term growth outlook.
This context matters because the financing initiative is being adjusted while demand for the underlying product remains exceptionally strong.
The Move Could Protect Nvidia From Regulatory Risk
By pausing selected revenue-sharing agreements, Nvidia may be trying to preserve the broader financing strategy while reducing the most controversial elements.
A revised program could potentially provide financing without giving Nvidia as much influence over customers’ commercial decisions.
Possible Revised Model
Current Model
Financing
GPU Sales
Capacity Backstop
Customer Restrictions
50% Revenue Share
↓
Regulatory Concerns
↓
Potential Revised Model
Financing
GPU Sales
Limited Capacity Support
↓
Lower Customer Control
↓
Reduced Regulatory Risk
↓
Continued Ecosystem Expansion
Nvidia’s spokesperson said the model is continuing to evolve, leaving open the possibility of such a restructuring.
AI Cloud Companies Could Face More Financing Uncertainty
For smaller cloud providers, the pause could complicate plans to secure capital for Nvidia-powered infrastructure.
These businesses typically need financing before their revenue base becomes large enough to support major expansion.
If Nvidia reduces its role as a financing partner, cloud companies may have to rely more heavily on banks, private credit, infrastructure funds and other investors.
Alternative Financing Sources
| Financing Source | Potential Role |
|---|---|
| Nvidia | Chip-linked financing |
| Commercial banks | Traditional debt |
| Private credit | Flexible infrastructure financing |
| Infrastructure funds | Long-term capital |
| Equity investors | Growth capital |
| Customer prepayments | Demand-backed funding |
| Leasing | Reduce upfront hardware cost |
Some AI cloud operators have already used combinations of these structures to fund expansion.
IREN, for example, has secured billions through customer prepayments, convertible notes, GPU leasing and GPU financing while planning a major expansion of its Nvidia-powered capacity.
The Revenue-Sharing Model Could Return In Another Form
The Reuters report makes clear that Nvidia has not permanently abandoned the concept.
The company could revise the program or fold it into another financing initiative.
That means the current pause may be better understood as a redesign phase rather than a strategic retreat from AI cloud financing.
Nvidia still has a strong incentive to support the expansion of the ecosystem around its chips.
More cloud capacity means:
More GPUs Installed
↓
More AI Customers
↓
More Compute Usage
↓
More Nvidia Chip Demand
The challenge is accomplishing that without creating regulatory concerns or excessive financial exposure.
Nvidia Is Taking On More Infrastructure Risk
Nvidia’s strategy increasingly extends beyond manufacturing and selling processors.
The company is providing capital, making investments, guaranteeing infrastructure obligations and supporting customer financing.
That can accelerate growth, but it also changes Nvidia’s risk profile.
Nvidia’s Expanding Role
| Traditional Nvidia | Emerging Nvidia |
|---|---|
| Chip designer | Chip designer |
| Hardware supplier | Hardware supplier |
| Software platform | Software platform |
| Networking provider | Networking provider |
| — | Investor |
| — | Financing partner |
| — | Capacity backstop |
| — | Infrastructure guarantor |
| — | Revenue-sharing partner |
This evolution is one of the most important strategic developments in the AI infrastructure market.
The Bigger Picture
Nvidia’s decision to pause some revenue-sharing deals shows the limits of its increasingly ambitious effort to finance and shape the AI infrastructure ecosystem. The company wants smaller cloud providers to purchase more Nvidia GPUs, build capacity and rent it to AI customers, but the proposed structure also gives Nvidia a financial interest in the downstream cloud business. That creates both a powerful growth mechanism and potential regulatory complications.
The pause does not appear to signal weaker AI-chip demand. Nvidia’s latest quarterly results showed $96.22 billion of revenue and approximately $89 billion from data centers, while the company expects roughly $108 billion of revenue in the next quarter. Instead, the development appears focused on how Nvidia can support its ecosystem without taking on excessive regulatory, financial or competitive risk.
Looking Ahead
The next step will be whether Nvidia restructures the revenue-sharing program or incorporates its financing features into another initiative. A revised model could preserve access to capital for smaller AI cloud providers while limiting Nvidia’s ability to dictate customer relationships, capacity allocation and downstream business decisions. The outcome will be closely watched by regulators, investors and cloud operators because it could establish a template for how dominant semiconductor companies finance the infrastructure built around their products.
For AI cloud companies, the pause reinforces the importance of diversifying their funding sources rather than relying solely on Nvidia-linked financing. Demand for AI compute remains extremely strong, but the economics of building thousands of GPUs into profitable cloud capacity are becoming more complex. Nvidia’s ability to balance chip sales, ecosystem expansion, financing support and regulatory constraints could ultimately determine how much influence it has over the next generation of AI infrastructure.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
