UPDATED 18:48 EDT / FEBRUARY 21 2024

INFRA

Nvidia’s data center GPU sales grow by a stunning 409% on huge demand for AI chips

Updated with Thursday stock price:

Nvidia Corp. once again crushed Wall Street’s expectations as it delivered its fourth-quarter financial results, sending its stock higher in extended trading.

The company also forecast revenue for the current quarter that came in well ahead of the Street’s forecasts, underscoring that there’s no end in sight to the technology industry’s seemingly insatiable thirst for artificial intelligence.

The company reported earnings before certain costs such as stock compensation of $5.16 per share, easily beating the analysts’ consensus estimate of $4.64 per share. Revenue for the period rose by an astonishing 265% to $22.1 billion, ahead of the $20.62 billion target, thanks in part to what the company said were strong sales of its “Hopper” graphics processing units, such as the H100 GPU.

All told, the company reported net income of $12.29 billion during the quarter, representing growth of 769% from the year ago quarter, when it delivered a profit of just $1.41 billion.

Looking to the current quarter, Nvidia said it’s expecting to deliver $24 billion in sales, versus the Street’s forecast of $22.17 billion in revenue.

The Street was once again thrilled, and Nvidia’s fast-growing stock made yet more gains, rising by more than 7% in the after-hours trading session. Update: Shares rose about 15% Thursday, giving it a market capitalization approaching $2 trillion.

Nvidia has become the primary beneficiary of the enormous demand for computing power to support AI projects, which have become a must for almost every enterprise these days thanks to the buzz around generative AI models such as ChatGPT. In a conference call, Nvidia Chief Executive Jensen Huang (pictured) attempted to address fears raised by some investors and analysts that the company might not be able to keep pace with the enormous demand for its GPUs, which are essential for training AI and powering workloads in production.

Huang insisted that the “conditions are excellent for continued growth” throughout fiscal 2025 and beyond. He explained that the company believes demand for its GPUs will remain elevated due to the interest in generative AI and also an industry-wide shift from central processing units to the company’s accelerator chips.

In a press release, the company said the demand for its technology was driven by both enterprise software and consumer internet applications across multiple industries, including automotive, healthcare and financial services.

Sales related to AI workloads are reported in Nvidia’s Data Center segment, which now accounts for the bulk of the company’s revenue. The unit reported revenue of $18.4 billion, up by 409% from a year earlier. Within that segment, more than half of all sales were attributed to large cloud infrastructure providers such as Amazon Web Services Inc.

Ever since OpenAI launched ChatGPT in November 2022, Nvidia has seen huge demand for its GPUs from technology giants such as Alphabet Inc. and Microsoft Corp., as well as generative AI startups such as Anthropic PMC and Cohere Inc., as they race with one another to launch increasingly powerful large language models that can perform more tasks. Generative AI models require massive amounts of computing power.

Last month, for instance, Meta Platforms Inc. CEO Mark Zuckerberg revealed that his company intends to buy more than 350,000 of Nvidia’s H100 GPUs by the end of the year, a statement that suggests it will spend billions of dollars on the chipmaker’s products.

Over the last three quarters, Nvidia’s data center business has grown enormously to become the clear market leader in the AI computing industry.

On the conference call, Nvidia Chief Financial Officer Colette Kress said its data center revenue would have been even higher if not for the new restrictions placed by the U.S. government on China. The company has been prohibited from shipping its most powerful products to Chinese customers over fears that its technology could end up in the hands of the country’s military.

The demand being placed on the data center business is such that its GPUs remain in short supply, and though Kress insisted the situation has improved, she admitted the company is still struggling to fulfill all of its customer’s orders.

“We are delighted that supply of Hopper architecture products is improving,” Kress told analysts. “Demand for Hopper remains very strong. We can expect our next-generation products to be supply constrained as demand far exceeds supply.”

Third Bridge analyst Lucas Keh said the restrictions on China and the supply concerns continue to be the biggest threats to Nvidia’s growth prospects. For instance, he said, Taiwan Semiconductor Manufacturing Co. is now close to hitting its capacity limit for the silicon wafers used by the company’s most advanced GPUs. “Investors should keep an eye out for Nvidia’s plans to diversify its supply base, and how quickly it will be able to do so,” Keh said.

The incredible rise of Nvidia’s fortunes in recent years represents the technology industry’s best example of a rags to riches story, for it was only about five years ago that the company’s future was looking extremely bleak, as none of the major cloud providers would embrace its GPU platforms, said Holger Mueller of Constellation Research Inc. Instead, the cloud vendors were betting on their own customized silicon, leaving Nvidia to ponder a future of shrinking on-premises-based growth.

As such, Nvidia spread its bets with its focus on AI, creating a diversification strategy that is now paying off handsomely, Mueller explained. “For a company in the double-digit billion dollar revenue range to double its sales is unheard of an unseen in the tech industry,” the analyst continued. “Despite this, there is no one really challenging Nvidia in AI yet, as AMD and Intel are barely blips on its radar with their first, nascent offerings. Going forward, the most urgent tasks for Jensen Huang and team are to keep fostering and building out an ecosystem, stay on top of the supply chain challenges and keep regulators happy. Do that and they’ll keep on winning.”

Nvidia’s gaming business, which sells GPUs for laptops, PCs and consoles, delivered revenue of $2.87 billion, up by 56% from a year earlier. The division used to be the company’s biggest revenue generator until the AI craze took off.

Elsewhere, the company reported that its automotive unit’s revenue declined 4% from a year earlier, to just $281 million, while revenue in the “OEM and other” segment, which includes chips for mining cryptocurrency, rose by 7%, to $90 million. Lastly, the professional applications unit saw sales rise by 105%, to $463 million.

Photo: Nvidia

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU