Nvidia pulls back from cloud services to double down on role as AI industry’s supplier

Instead of pushing cloud services straight to customers, Nvidia is turning DGX Cloud inward by using it for its own needs and to help out its partners. The move signals a clear shift in strategy for a company that briefly explored competing with the same cloud providers that rely heavily on its chips.

According to reports, Nvidia has folded the DGX Cloud team into its core engineering organization and reassigned leadership tied to the cloud business. As a result, DGX Cloud will now focus on supporting Nvidia’s own engineers who build and test artificial intelligence models. Nvidia just changed course from what it said back in 2023, when it hinted at going head-to-head with the big cloud providers.

Back then, Nvidia talked up DGX Cloud as a high-end service, promising top performance by keeping a close grip on the hardware. However, the company entered a market already dominated by Amazon Web Services, Google Cloud, and Microsoft Azure. Therefore, gaining attention and customers proved difficult from the start. These providers also represent Nvidia’s largest buyers, which added another layer of tension to the effort.

Technical hurdles further complicated the initiative. Nvidia ran DGX Cloud across infrastructure leased from multiple cloud operators. Consequently, teams struggled to troubleshoot issues consistently because fixes in one environment rarely applied to others. Meanwhile, pricing pressure intensified.

More importantly, Nvidia appeared increasingly cautious about damaging its core relationships. Cloud providers account for a large share of Nvidia’s revenue through GPU purchases. Because of that, executives reportedly worried that expanding DGX Cloud could encourage partners to accelerate development of their own AI chips. Some providers declined to participate in the service altogether, while others showed limited engagement.

Nvidia also absorbed its DGX Cloud Lepton marketplace into internal operations after it failed to gain momentum. Nvidia isn’t doubling down on services right now. Instead, the company’s guarding its lead in AI hardware. It still owns most of the AI chip market, and it’s pouring major resources into cloud infrastructure—but really, that’s about boosting its own research and development.

WIth Nvidia’s stepping back from chasing cloud market share, it’s making a clear choice: stay right at the heart of the AI supply chain, rather than go head-to-head with its biggest customers. That move says a lot—it’s all about holding onto influence in the broader ecosystem, not just grabbing every piece of the pie.