Shulou(Shulou.com)11/24 Report--

CTOnews.com, Sept. 12 (Xinhua)-- Nvidia is trying to use its dominant position in AI hardware to get a piece of cloud software. The move also allowed Nvidia to compete with its customers, the traditional cloud service providers that bought its AI chips, which also resented some customers.

Due to a series of effects after OpenAI released ChatGPT last year, especially the large number of AI developers springing up after a spring rain, there has been a surge in demand for servers with Nvidia AI chips in various industries, and these cloud service providers soon find that their own computing power can no longer meet the needs of the market.

At this delicate moment, Nvidia naturally sees business opportunities.

According to theinformation, Nvidia made an unusual proposal last year to Amazon Web Services,AWS and other cloud computing providers, which had been the biggest buyers of Nvidia's AI chips.

To put it simply, Nvidia wants to rent customers' servers with Nvidia chips. This allows it to sublease customers' servers to AI software developers, which naturally include some of the world's largest cloud developers.

As Nvidia's chips continue to grow, such as the new generation of H100 chips it holds, it happens to be the most urgent need for these cloud providers at that time, some cloud service providers have to bow.

Microsoft, Google and Oracle all agreed to Nvidia's proposal, but AWS did not because it feared that Nvidia's entry would threaten their position in the market, according to a person familiar with the decision.

According to the latest quarterly results, Nvidia's revenue for the current quarter was $13.51 billion, up 88 per cent from the previous quarter and 101 per cent from a year earlier.

After a blowout growth in May, Nvidia has become the first semiconductor company in history to have a market capitalization of $1 trillion. Since then, investors have been looking for evidence that it will maintain strong growth in the second quarter.

In March, Huang Renxun, chief executive of Nvidia, announced a plan to open supercomputers used to develop artificial intelligence technologies such as ChatGPT to commercial customers. Almost any company can use these powerful but expensive devices through cloud leasing.

Of course, Nvidia's rent is not cheap, with eight A100 or H100 flagship chips priced at $37000 a month (CTOnews.com Note: currently about 270000 yuan). However, opening the supercomputer to a wider range of business customers will accelerate the development of artificial intelligence. The popularity of artificial intelligence technology in the past has pushed Nvidia's shares up 77% this year, with a market capitalization of about five times that of longtime rival intel.

Take another look at Amazon AWS, which did not receive assistance from Nvidia until July to launch a new instance of Amazon Elastic Compute Cloud (EC2) P5 based on the Nvidia H100 Tensor Core GPU. According to reports, the service allows users to easily extend generative AI, high performance computing (HPC) and other applications through browsers.

At Nvidia's annual developer conference, Huang Renxun said the company was working with partners such as Oracle to provide cloud access to Nvidia's DGX supercomputer. The supercomputer integrates as many as 32000 Nvidia chips. Through this service, anyone who has a web browser can log on to the supercomputer.

Related readings:

"to promote cloud computing and generative AI, Nvidia H100 GPU has been opened to Amazon AWS cloud service users"

Nvidia launches supercomputer cloud rental service: a package with eight A100 or H100 flagship chips costs $37000 a month.

"Nvidia has made a lot of money. Oracle will spend billions of dollars on GPU AI cloud service this year."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.