Components & Peripherals News

Nvidia T4, V100 Sales Push Data Center To Double-Digit Growth

Dylan Martin

'I believe that in the future the data center will all be accelerated,' Nvidia CEO Jensen Huang says of his belief that GPUs will drive acceleration for most server workloads in the future.


Nvidia credited record sales of its Tesla V100 training GPUs and T4 inference GPUs for moving its data center business from recovery mode to double-digit growth.

The Santa Clara, Calif.-based company announced on Thursday that its data center revenue reached a record $968 million in the fourth quarter of its 2020 fiscal year, a 43 percent increase over the same period last year, which marked a continued recovery from a contraction earlier in the year.

[Related: Intel Axes Nervana AI Chips In Favor Of Habana Labs]

Jensen Huang, Nvidia's CEO and founder, said V100 and T4 sales were driven by the need for GPUs to accelerate artificial intelligence and deep learning workloads. According to Nvidia CFO Collette Kress, T4 sales in the fourth quarter were four times greater than the same period last year.

"The primary driver for our growth is AI," Huang said.

The chipmaker's stock price was up as much as 6.40 percent in after-hours trading Thursday after reporting results for the fourth quarter that surpassed Wall Street's expectations.

Nvidia's fourth-quarter revenue was $3.1 billion, a 40.7 percent increase from the same period last year and $130 million higher than what analysts had hoped to see. The company's net earnings were $1.89 cents per share, beating analyst estimates by 22 cents.

Huang said there are "four fundamental dynamics" behind Nvidia's AI growth: breakthroughs being made in AI models for things like natural language understanding and conversational AI, AI deployments moving into production with inference workloads, growing adoption from hyperscale and vertical customers, and the emergence of processing AI at the edge.

The CEO said he is bullish on the chipmaker's long-term prospects in the data center because he believes GPUs will become key to every workload in the future.

"I believe that in the future the data center will all be accelerated," Huang said.

Echoing industry observers who say, "Moore's law is dead," Huang said he believes the prominence of CPUs made by companies like Intel and AMD will fade as crucial parts of compute engines.

"CPU scaling has really slowed. There's no two ways about it. It's not a marketing thing. It's a physics thing," he said, adding that Dennard scaling, the law that states power density remains constant as transistors get smaller, has come to an end.

Nvidia projected that revenue for the first quarter of its fiscal 2021 year will be $3 billion, plus or minus 2 percent. The company said it had lowered its forecast for the quarter by $100 million due to the coronavirus outbreak, which is expected to impact the company's gaming and data center units equally.

"At this time, that was our best estimate," Kress said.

In response to an analyst question about supply, Huang said while Nvidia is "in pretty good shape," the company "won't have ample supply."

"We just need to watch it very closely," he said, adding that Nvidia is working closely with partners on forecasting. "All of us have to do a better job forecasting."

Learn More: CPUs-GPUs
Dylan Martin

Dylan Martin is a senior editor at CRN covering the semiconductor, PC, mobile device, and IoT beats. He has distinguished his coverage of the semiconductor industry thanks to insightful interviews with CEOs and top executives; scoops and exclusives about product, strategy and personnel changes; and analyses that dig into the why behind the news.   He can be reached at

Sponsored Post