Nvidia pushes further into cloud with GPU marketplace
Summary
The chip giant has created DGX Cloud Lepton, a new service that will make its AI chips directly available to developers across a variety of cloud platforms.Nvidia is a relative newcomer to the cloud-computing game, but it’s quickly gaining momentum. The semiconductor giant on Monday announced a service that makes its AI chips available on a variety of cloud platforms—widening access beyond the major cloud providers.
The service, called DGX Cloud Lepton, is designed to link artificial intelligence developers with Nvidia’s network of cloud providers, which provide access to its graphics processing units, or GPUs. Some of Nvidia’s cloud provider partners include CoreWeave, Lambda and Crusoe.
“Nvidia DGX Cloud Lepton connects our network of global GPU cloud providers with AI developers," said Jensen Huang, chief executive of Nvidia in a statement. The news was announced at the Computex conference in Taiwan.
Leading cloud service providers are expected to also participate, Nvidia said. The move makes its chips more widely accessible to developers of all kinds—not just those who have relationships with those tech giants, analysts say.
“We saw that there was a lot of friction in the system for AI developers, whether they’re researchers or in an enterprise, to find and access computing resources," said Alexis Bjorlin, Nvidia’s vice president of the DGX Cloud unit.
DGX Cloud Lepton is a one-stop AI platform with a marketplace of GPU cloud vendors that developers can pick from to train and use their AI models, Nvidia said.
Since the AI boom kicked off in late 2022, Nvidia’s GPUs have been a hot commodity. Cloud providers have been racing to gobble up chips to support both their customers and their own internal AI efforts.
But at any given time, cloud providers—including smaller players like CoreWeave—might have GPUs that aren’t being used. That’s where Lepton comes in, Bjorlin said, because it’s a way for those providers to tell developers they have excess computing for AI.
“This is Nvidia’s way to kind of be an aggregator of GPUs across clouds," said Ben Bajarin, chief executive and principal analyst of market-research firm Creative Strategies.
Nvidia will be reaching developers directly, rather than going through its cloud-provider partners. That kind of direct outreach furthers Nvidia’s aim of building its business with enterprises, and not just AI labs, said International Data Corp. analyst Mario Morales.
The other benefit of working directly with the chip giant is that developers can choose which AI cloud provider to work with, or choose to work with multiple cloud providers, Nvidia’s Bjorlin said.
“It is up to the developer to choose," Bjorlin added. “Nvidia is not intersecting in that pathway."
Write to Belle Lin at belle.lin@wsj.com
topics
