Nvidia’s coming year will be a turbulent one

Nvidia’s breakneck growth over the past 18 months has definitely put the company on top—both of the burgeoning artificial-intelligence market and the stock market. The chip maker’s fiscal third-quarter report Wednesday afternoon showed annual sales crossing the $100 billion mark for the first time ever—more than double what it was generating a year ago. And the company’s $3.6 trillion market capitalization is more than $100 billion higher than that of Apple.

The last point is particularly fitting since hype over Nvidia’s data center chips, known as GPUs, now eclipses what once greeted new iPhone launches. But there is a downside too: Fully living up to the hopes that have made it the world’s most valuable company can still be a stretch. Nvidia’s third-quarter results beat Wall Street’s targets across the board, as did the company’s forecast for the current quarter. But the stock still slipped about 2% in after-hours trading since the projection beat Wall Street’s consensus estimate by the lowest margin since the company’s AI business began taking off in May of last year.

Investors will likely shake it off. The company’s previous report in August sparked an even bigger selloff, yet the shares have rebounded 24% since. But such volatility will likely be the new normal for Nvidia as it heads into a new year with a hotly anticipated new product line that is also so complex it will be constrained by supply and production challenges. There is also growing uncertainty about more tariffs and other potential roadblocks from the incoming Trump administration that could further hobble Nvidia’s sales in China, where it is already restricted from selling its most sophisticated chips.

And finally, how do investors value an Nvidia that is still growing quickly, but not as quickly as it used to be? Revenue jumped 94% year over year to $35.1 billion in the quarter ended in October, which is the first time in five quarters that Nvidia’s growth rate hasn’t cracked triple digits. The company projected revenue growing 70% year over year to $37.5 billion in the current period.

That projection includes the first volume shipping of the new Blackwell family of AI systems that Nvidia unveiled in March. But the sequential change of just $2.4 billion from the just-ended quarter leaves some uncertainty as to how quickly Blackwell will be able to generate a substantial amount of new revenue. The timing of the rollout was the dominant topic during the question-and-answer session of Nvidia’s earnings call Wednesday.

“Blackwell demand is very strong," Nvidia Chief Executive Officer Jensen Huang said on the call, while later reminding listeners, “We guide one quarter at a time."

It needs to be big. Analysts expect the Blackwell family to generate $62.6 billion in sales for Nvidia’s fiscal year ending in January of 2026 and more than $97 billion in the following fiscal year, according to consensus estimates from Visible Alpha. Note that even $62 billion is greater than the entire annual revenue of every other chip company in the world save for Taiwan Semiconductor Manufacturing, Nvidia’s main production partner.

Such lofty targets are attainable—especially if Nvidia’s largest customers keep spending the way they’ve promised to in their own most recent quarterly calls. Deep-pocketed tech giants Microsoft, Amazon.com, Meta Platforms and Google-parent Alphabet are projected to drop more than $242 billion combined this year on capital spending and an additional $285 billion next year, according to Visible Alpha estimates.

But that investment will depend on how quickly their own customers adopt generative-AI services, as well as their ability to build and power the types of data centers capable of fully running Nvidia’s demanding systems. Much has gone right for Nvidia over the past year. More needs to keep going right for the next one.

Write to Dan Gallagher at dan.gallagher@wsj.com