Original author: Dong Jing

Original source: Wall Street Insight

NVIDIA is transforming its powerful balance sheet into market leverage by providing financial backing to emerging cloud service providers in exchange for revenue sharing, quietly evolving from a chip seller to the "central bank" of the AI computing ecosystem.

On July 1, according to technology media The Information, NVIDIA is providing financial guarantees to young cloud service providers renting and selling their GPUs—if these companies are unable to find enough AI developers to rent computing power, NVIDIA will lease back their unsold GPU capacity at an agreed price.

In exchange, NVIDIA will take a certain percentage of revenue from these cloud service providers, and the revenue share percentage will gradually decrease over the contract period. GPU cloud service firms Firmus and Sharon AI have already participated in the project, and three executives with business dealings with NVIDIA have confirmed the related arrangements.

NVIDIA announced on its official website on July 1 the launch of a new business model combining revenue sharing and credit support, allowing AI cloud vendors to procure NVIDIA infrastructure without fully bearing the upfront capital expenditures, and to provide computing power services to downstream AI-native firms, model developers, and enterprise customers.

Reports indicate that this project is referred to internally within NVIDIA by some as the "AI Compute Partnership." A spokesperson for NVIDIA has also confirmed the existence of the project. This move marks a significant strategic shift for NVIDIA:

On one hand, expanding the customer base by lowering the financing threshold for emerging cloud providers, and on the other hand, directly participating in the profit distribution of the downstream computing power market through revenue sharing, further extending its control over the AI industry chain downstream.

Model Transition: From Selling Chips to Sharing Cloud Revenue

According to NVIDIA's official press release, NVIDIA will gain additional income from cloud service revenue beyond standard product revenues, thus forming a recurring revenue stream linked to usage. The core intention of this model is to break down the longstanding financing barriers that have restricted startup AI companies from obtaining large-scale computing power.

NVIDIA positions this framework as the "DSX AI Factory" model, aimed at AI service scenarios that require cross-regional continuous operation, high utilization, and multi-tenant accelerated computing.

Sharon AI and Firmus are the first cloud vendors 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 park on Batam Island in Indonesia, expected to scale up to 360 megawatts, equipped with up to 170,000 NVIDIA GPUs. These deployments directly reflect NVIDIA's latest progress in converting computing power demand into feasible, financeable infrastructure.

NVIDIA points out that emerging AI companies have historically faced serious limitations in obtaining capital-intensive infrastructure—signing long-term commitment contracts often does not suffice to facilitate financing for computing power procurement. This means that many AI-native companies, model developers, and inference service providers have had to face long waits when expanding their computing capabilities: site selection, power procurement, construction, hardware debugging—each step can take months or even longer.

The promise of the new model is: by realigning the economic structure, to enable the aforementioned groups to more quickly access full-stack accelerated computing capabilities, without waiting for traditional infrastructure construction cycles to complete.

Guarantee Logic: Solving the Core Financing Problem of GPUs

According to reports, GPUs are typically the most expensive component in AI data centers. For buyers of chips with low credit ratings, securing adequate loans is a hurdle in itself.

A data center executive commented that NVIDIA's transactions of this nature are "killing two birds with one stone." He explained that if NVIDIA only backs the lease for data center facilities, "you still face the problem of 'how to finance the GPUs'"; but if NVIDIA commits to cover the unsold computing power in facilities, "the financing issue for GPUs is resolved, as is the financing issue for the data center."

In other words, NVIDIA's guarantee essentially acts as a credit enhancement tool, enabling emerging cloud service providers, who originally had difficulty obtaining bank loans, to leverage larger capital amounts and accelerate data center construction.

Strategic Intent: Breaking the Monopoly Pattern of Major Clients

NVIDIA has launched this series of measures against a clear strategic backdrop. Currently, a few large cloud service providers, including Amazon, Microsoft, SpaceX, Oracle, Meta, and Google, have purchased most of NVIDIA's chip capacity. However, several of these companies are developing competitive AI chips in-house, which poses a potential threat to NVIDIA.

To reduce dependency on these giant clients, NVIDIA has continuously supported a group of emerging GPU cloud service providers represented by CoreWeave over the past few years. The "AI Compute Partnership" is a continuation and deepening of this strategy.

According to previous reports from The Information, NVIDIA has been negotiating to provide financial guarantees for OpenAI to rent a large data center in Ohio, which, if fully built at current prices for chips, labor, power, and other materials, could cost as much as $500 billion.

Financial Investment: From Equity Investment to Capacity Guarantees

NVIDIA's financial investment in this direction has already been 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, involving firms like CoreWeave and Lambda, with total transaction amounts reaching several billion dollars. According to previous reports from The Information, NVIDIA's own researchers have also used GPU servers leased from Lambda.

In terms of capacity guarantees, NVIDIA began pushing related transactions last fall. In September 2024, NVIDIA promised to buy all of CoreWeave's unsold capacity until 2032 if it cannot find tenants, at that time the contract was valued at $6.3 billion. This move effectively alleviated investor concerns about CoreWeave's high-leverage business model, pushing its stock price up nearly 30% in the following week.

Overall, NVIDIA is building a multi-layered mechanism of interests binding: equity investments, capacity leasing, lease guarantees, and now revenue sharing. Each layer deepens the financial ties between NVIDIA and downstream cloud service providers, also enabling NVIDIA to directly share the incremental revenue from the commercialization of AI computing power, beyond chip sales.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。