Components & Peripherals News
Nvidia: Data Center Bright Spot In Relatively Dark Third Quarter
Joseph F. Kovar
‘Over the years, our rich ecosystem and our software stack has been integrated to developers and startups of all kinds. But more than ever, we‘re at the tipping point of clouds. And that’s fantastic, because if we could get Nvidia’s architecture and our full stack in every single cloud, we could reach more customers more quickly,’ says Nvidia Founder and CEO Jensen Huang.
Massive growth in Nvidia’s data center business was not enough to overcome the macroeconomic headwinds that pushed back on the performance chip, hardware, and software developer’s fiscal 2023 third quarter financial results.
A 31-percent rise in Nvidia’s data center sales, along with 86-percent growth in its much smaller automotive chip and software business, were not enough to overcome the impact from a 51-percent drop in gaming sales and 65-percent drop in the company’s professional visualization revenue.
Colette Kress, Nvidia executive vice president and chief financial officer, told analysts Wednesday during her prepared remarks on the company’s quarterly financial analyst call that while Santa Clara, Calif.-based Nvidia was able to start shipping its new flagship H100 data center GPUs, based on its Hopper architecture, the company faced U.S. government export restrictions related to its A100 and H100 products to China.
Nvidia also faced lower PC demand and a shift in Ethereum cryptocurrency transactions that sent many used GPUs to the aftermarket, impacting the sale of GPUs and related products.
When asked by an analyst about the factors impacting the growth of Nvidia’s data center business, Founder and CEO Jenson Huang responded that there are two fundamental data center dynamics happening.
The first has to do with general purpose computing no longer scaling, making acceleration the key to achieving the necessary level of cost efficiency scale and energy efficiency scale so that users can continue to increase workloads while saving money and power, Huang (pictured) said.
The second dynamic is AI, an area where Nvidia is seeing surging demand in some very important sectors and related AI breakthroughs such as deep recommender systems, which are essential to recommending the best content or product for users using a mobile device or a voice-activated computer system, Huang said.
“You need to really understand the context of the person making the request and make the appropriate recommendation to them,” he said. This involves large language models, which have evolved over time to encompass not only different human languages but also vertical-specific languages, Huang said.
Another growing segment of the AI space is generative AI, or the ability of AI to help created an actual product, as opposed to perception AI which is aimed at understanding context, Huang said. “This is now the beginning of the era of generative AI,” he said. “You probably see it all over all over the place, whether they‘re generating images or generating videos or generating text of all kinds.”
Huang, responding to an analyst question about the impact of hyperscale cloud providers on Nvidia’s business, said that Nvidia is making progress with several hyperscale cloud partners.
“You know that Nvidia has a rich ecosystem,” he said. “And over the years, our rich ecosystem and our software stack has been integrated to developers and startups of all kinds. But more than ever, we‘re at the tipping point of clouds. And that’s fantastic, because if we could get Nvidia’s architecture and our full stack in every single cloud, we could reach more customers more quickly.”
For that reason, Nvidia has recently unveiled several cloud-focused partnerships and collaborations, including one unveiled Wednesday with Microsoft aimed at scaling AI to meet the needs of startups clamoring for large installations of Nvidia GPUs for large language model training, and for building startups and scaling AI to the enterprise internet service providers, Huang said.
The company is also now working with Oracle, Rescale, and other cloud service providers, he said.
“Every company we‘re talking to would like to have the agility, the scale, and the flexibility of clouds. … The goal of all of this is to move the Nvidia platform full-stack software into the cloud so that we can engage customers much more quickly and customers could engage our software. If they would like to use it in the cloud, it’s per GPU instance-hour. If they would like to utilize our software on-prem, they could do it through a software license and subscription. And so in both cases we now have software available practically everywhere you would like to engage it.”
For its fiscal 2023 third quarter, ended October 30, Nvidia reported revenue of $5.93 billion, down 17 percent from the $7.10 billion the company reported for its third fiscal quarter 2022.
The total revenue figure beat analyst expectations by $110 million, according to Seeking Alpha.
The results included data center revenue of $3.83 billion, up 31 percent year over year; gaming revenue of $1.57 billion, down 51 percent; professional visualization revenue of $200 million, down 65 percent; and automotive and embedded revenue of $251 million, up 86 percent.
Nvidia reported GAAP net income of $680 million or 27 cents per share, a huge plunge from last year’s $2.46 billion or 97 cents per share. On a non-GAAP basis, Nvidia reported net income of $1.46 billion or 58 cents per share, down significantly from last year’s $2.97 billion or $1.17 per share earnings.
Analysts were expecting non-GAAP earnings of $1.29 per share, according to Seeking Alpha.
Looking forward, Nvidia expects fiscal fourth quarter revenue of $6.00 billion plus or minus 2.0 percent. Analyst consensus was $6.14 billion, according to Seeking Alpha. That compares to fourth fiscal quarter 2022 revenue of $7.6 billion, which was itself up 53 percent compared to the same period in fiscal 2021.