• Nvidia has reduced efforts to promote its DGX Cloud service, stepping back from direct competition with major cloud providers like Amazon Web Services.
  • The strategic shift reinforces Nvidia's focus on its core competency as a hardware supplier to cloud giants, rather than a direct rival in hosting enterprise AI workloads.
  • The move coincides with AWS's rollout of new servers powered by Nvidia's latest Blackwell GPUs, highlighting the complex partnership-competition dynamic between the two firms.

Nvidia Corp. has significantly dialed back its efforts to attract businesses to its DGX Cloud service, according to a person with direct knowledge of the matter, marking a strategic retreat from its ambition to compete directly with hyperscalers like Amazon Web Services Inc. The initiative, which offered fully managed high-performance AI compute via the cloud, has seen its promotional push lessened as the chipmaking behemoth refocuses on its primary role as a hardware supplier.

The decision underscores the formidable challenge of competing with entrenched cloud infrastructure leaders. Rather than battling for cloud market share, Nvidia appears to be doubling down on its immensely profitable core business: designing and selling the advanced GPUs that power the AI revolution across all platforms. This pivot comes at a time when demand for its processors, like the H100 and the newly announced Blackwell Superchips, continues to far outstrip supply, making the resource-intensive task of building a competing cloud service less attractive.

This strategic realignment was not explicitly announced but is evident in the company's shifting focus. Efforts to secure major enterprise contracts for DGX Cloud have been deprioritized, the person said. Nvidia did not immediately respond to a request for comment on the shift in strategy.

The timing is particularly notable given the recent announcement from AWS regarding the general availability of its P6e-GB200 UltraServers, which are accelerated by Nvidia’s very own Blackwell chips. This partnership exemplifies the nuanced relationship between the two companies: Nvidia supplies the critical hardware that enables AWS's cutting-edge AI services, even as it had previously hoped to host similar workloads on its own platform.

For enterprise customers and AI developers, Nvidia's pullback means the landscape for accessing high-performance AI compute remains concentrated among the largest cloud providers. While DGX Cloud offered a direct path to Nvidia's latest technology, its scaling back may simplify decisions for businesses that were evaluating it against established offerings from AWS, Microsoft Azure, and Google Cloud Platform. The move ultimately reinforces the position of hyperscalers as the primary gateways to AI infrastructure.

Analysts suggest this is a pragmatic move by Nvidia. The capital expenditure required to build a global cloud infrastructure capable of competing with AWS is astronomical, and the margins are likely far thinner than those enjoyed from its near-monopoly on high-end AI training chips. By supplying its hardware to all major clouds, Nvidia captures value from the entire ecosystem without bearing the operational burden and competitive risk of running its own.