Nvidia eases restrictions for cloud customers and expands chip rental program

The program was launched in July by NVIDIA to make it easier for cloud providers to access coveted graphics processors needed for training and running large AI models. Nvidia itself does not own significant cloud resources and therefore relies on external partners to deliver its chips to end customers. The financing model is intended to enable smaller providers to enter the market while securing Nvidia's influence across the entire AI value chain.

Background: Chip shortages and Nvidia's powerful position

According to the Wall Street Journal's account, Nvidia is responding to resistance from the industry with this loosening. Several AI cloud partners had insisted on being able to freely choose their own customers without having to accommodate a list drawn up by the chipmaker. Industry observers interpreted the move as an acknowledgment that the original requirements were slowing the pace of the program's expansion.

The backdrop is the enormous demand for high-performance chips that has emerged since the boom in generative AI systems. Nvidia is considered one of the biggest beneficiaries of this trend and has reported record revenues in recent fiscal years. However, demand far exceeds available capacity, which is why the company is looking for ways to distribute its chips more broadly without jeopardizing its high-margin direct business with hyperscalers such as Microsoft, Google, or Amazon.

It initially remained unclear how far-reaching the changes actually are. The Wall Street Journal wrote only that Nvidia had loosened its requirements for end-customer selection. A company spokesperson declined to comment when contacted. No further details were provided either about the conditions for AI cloud partners, such as minimum purchase quantities or rental terms.

Reactions from the industry

The move is nevertheless seen in the industry as a signal that Nvidia is placing greater emphasis on growing its cloud ecosystem. Industry figures such as Raj Mirpuri, founder of the AI cloud provider Vultr, had repeatedly stressed in recent months that smaller and mid-sized providers need significantly better access to Nvidia's top-tier chips in order to compete with the major hyperscalers.

Implications for the AI cloud ecosystem

The revenue-sharing program through which Nvidia supports cloud providers using its chips is a central pillar of the company's strategy. It is intended both to secure additional revenue for Nvidia and to prevent competing chip architectures from gaining ground in the AI market. Industry observers see it as an attempt to maintain market dominance even when the company's own processors, such as the upcoming Rubin generation, are delivered with delays.

For AI cloud partners, the loosening means more negotiating leverage in customer acquisition in the short term. In the medium to long term, according to analysts' assessment, the program could also help to further increase the share of Nvidia chips in data centers outside the major hyperscalers, thereby strengthening the company's position across the entire AI ecosystem.