Nvidia’s AI Party Is Far From Over

Nvidia wasn’t priced for perfection. Investors expected it anyway—and nearly got it.

The chip maker at the forefront of the artificial-intelligence revolution managed to deliver a stronger-than-expected quarterly report and forecast on Wednesday afternoon. The numbers countered a growing worry among investors that a company whose market value has boomed by more than $1 trillion in just the last 12 months couldn’t possibly live up to the hype. Nvidia’s stock had slipped 9% by Wednesday’s closing bell from the all-time high it hit last week.

The stock clawed back much of that loss in after-hours trading following the results. Nvidia’s revenue for the fiscal fourth quarter ended January and its revenue forecast for the current quarter both beat Wall Street’s consensus estimates by 10%. Revenue for the data center segment that includes the accelerator chips designed for generative-AI computing hit $18.4 billion—five times its level from a year ago. Nvidia said during its conference call that it expects data center revenue to keep growing sequentially in the April quarter, when it expects total revenue to have tripled from the same period last year.

Nvidia’s long-established practice of projecting only one quarter out still could fuel questions about the longer-term trajectory of its AI business. Semiconductor markets are notoriously cyclical, and Nvidia’s data center business, which has long been selling chips for earlier forms of AI, has proven to be no exception. Growth in that unit has gyrated over the years as cloud computing giants sometimes take “digestion” periods following major purchases.

But such a swing doesn’t seem on the horizon anytime soon. Microsoft, Amazon, Google-parent Alphabet and Facebook-parent Meta Platforms all told investors during their most recent earrings calls that they plan to boost capital spending this year directly for the type of AI components that Nvidia provides. And on Wednesday’s call Nvidia Chief Financial Officer Colette Kress said “demand far exceeds supply” for the company’s current chips, and that next-generation products are also expected to be “supply constrained.” Nvidia is expected to share more about those products at its annual AI developers conference next month.

While Nvidia is now facing more competition from rivals like AMD and Intel—as well as chips designed in-house by its largest customers—the company hardly seems to be cutting prices to move its products. Adjusted operating income of nearly $14.8 billion for the recent quarter blew past Wall Street’s targets and produced an operating margin of 67%—Nvidia’s highest in at least a decade and far above the 43% margin averaged over the last eight quarters. Nvidia’s cost projections also indicated operating margins staying at that level in the current quarter ending in April.

Strong earnings growth also tempers worries over what otherwise would be a soaring valuation multiple. Nvidia was trading at 34 times projected earnings for this year at last week’s peak, which is actually 18% below the stock’s three-year average, according to FactSet data. The recent selloff puts Nvidia’s current valuation below nearly half the stocks in the PHLX Semiconductor Index, according to FactSet.

Nvidia’s biggest challenge in the near term might come down to managing the sort of expectations that have come with its rapid ascent. The unprecedented nature of the AI boom has produced a range of guesses as to where Nvidia’s business ultimately is heading. Analyst targets for the data center segment for the January quarter varied by more than $3.5 billion between the high and low end of estimates—twice the spread of the previous quarter, according to data from Visible Alpha.

Some options traders have even been betting on the stock reaching $1,300—nearly double its current price. That would equate to a market value well over $3 trillion, exceeding those of both Microsoft and Apple. Nvidia’s AI boom is no flash in the pan, but some speculators could still end up with a lump of fool’s gold.

Write to Dan Gallagher at [email protected]