Nvidia unveils DGX cloud, access to its supercomputers that power ChatGPT AI

"The iPhone moment of AI has started," says Jensen Huang, CEO of Nvidia.
Baba Tamim
NVIDIA CEO Jensen Huang during the presentation of the company’s cloud service announcement.
NVIDIA CEO Jensen Huang during the presentation of the company’s cloud service announcement.


Nvidia has collaborated with Oracle Cloud, Microsoft Azure, Google Cloud, and other cloud providers to make its A.I. supercomputers accessible via cloud services.

The idea is to make practically any firm able to borrow Nvidia's powerful and costly supercomputers used to create A.I. technologies like ChatGPT, according to the company's plan shared on Tuesday by CEO Jensen Huang.

"What we've done over the years with DGX is not just [create] a state-of-the-art supercomputer, but we've built a software stack that sits on top of it," said Manuvir Das, Nvidia's vice president of enterprise computing, during the presser.

"That turns this into turnkey training as a service. You just provide your job, point to your data set, and you hit go, and all of the orchestration and everything is taken care of."

"In DGX Cloud, now the same model is available on infrastructure hosted at various public clouds. It's the same interface, the same model for running your training," added Das. 

Nvidia DGX Cloud, the company's new offering, will allow enterprises to use a web browser to access Nvidia's DGX AI supercomputers and A.I. software immediately needed to train models for generative A.I. and other A.I. applications.

'iPhone moment of A.I.'

Offering access to a larger group of business customers could speed up the A.I. boom that has driven Nvidia shares up 77 percent this year, making it about five times more valuable than longtime rival Intel Corp.

The cost of the access, however, will be as high as $37,500 per month for eight of Nvidia's A100 or H100 flagship chips connected—it will be made available to a more extensive range of commercial clients.

"We are at the iPhone moment of A.I.,” Huang said in the virtual keynote address, referring to how Apple Inc. pioneered the smartphone market.

Moreover, Nvidia is teaming with AT&T Inc (T.N.) to improve truck dispatching, with quantum computing experts to accelerate software development, and with industry giant Taiwan Semiconductor Manufacturing Co (2330. T.W.) to accelerate chip development, added Huang.

The Santa Clara, California-based company also unveiled new hardware and software that would make supercomputer-created products like chatbots far less expensive to use daily.

The products "are years ahead of the competition," according to Hans Mosesmann, a semiconductors analyst at Rosenblatt Securities. 

"Nvidia's leadership on the software side of A.I. is not only monumental - it is accelerating," he said. 

Add Interesting Engineering to your Google News feed.
Add Interesting Engineering to your Google News feed.
message circleSHOW COMMENT (1)chevron
Job Board