Nvidia’s Profit Doubles as A.I. Chip Sales Soar

The company, which dominates the market for chips used to build artificial intelligence, expects another big jump in the current quarter.

Nvidia, which dominates the market for computer chips used to build artificial intelligence, has become a Wall Street sensation.

Over the past year, it has increased its market value by $2.5 trillion and dethroned Apple as the world’s most valuable tech company. But Nvidia’s success has raised questions about how long its run can last.

On Wednesday, Nvidia showed that the world’s seemingly insatiable appetite for A.I. technology could lift its business to new highs. The company, a bellwether for A.I., topped Wall Street’s expectations for the quarter, reporting that revenue jumped 94 percent from a year earlier and that profit increased 106 percent during the three months that ended in October.

Nvidia also projected that revenue in the current quarter would rise 70 percent from a year ago to $37.5 billion, as it began selling a more powerful A.I. chip called Blackwell. The forecast exceeded Wall Street’s prediction by about $500 million and suggested that customers were lining up to buy the new chip.

Revenue in the last quarter was $35.08 billion, surpassing the company’s estimate of $32.5 billion in August. Net income rose to $19.04 billion from $9.24 billion a year earlier, exceeding the quarterly profits of Amazon and Meta.

Shares in Nvidia fell 1 percent in after-hours trading, partly because its sales outlook was dampened by supply constraints on its new chip. The company said it would take several quarters to produce as many Blackwell chips as its customers wanted.

“The challenge that we have is how fast can we get that supply, getting ready, into the market this quarter,” said Colette Kress, Nvidia’s finance chief, during a call with analysts. “We’ll be back on track with more suppliers as we turn the corner into the new calendar year. We’re just going to be tight for this quarter.”

The results are the latest proof that companies are continuing to pour billions of dollars into generative A.I. technology. Nvidia’s chips, which can perform hundreds of tasks at the same time, are being purchased to power chatbots, write software code and discover new drugs. Those efforts have continued, even as some skeptics on Wall Street and in Silicon Valley question whether the financial return on those applications will justify their staggering costs.

Nvidia gained a dominant position in A.I. because its chief executive, Jensen Huang, bet that its semiconductors could be essential to building the technology. He spent years developing software and servers that would allow the chips, known as graphics processing units, or GPUs, to train systems to recognize images and predict words. The wager helped him corner 90 percent of the chip market.

Mr. Huang has become one of A.I.’s biggest proponents, saying its economic effect will be similar to the Industrial Revolution. He has traveled in recent weeks to India, Japan and Indonesia to encourage the countries and their companies to invest in building data centers.

“The computer industry has fundamentally changed,” Mr. Huang said in Japan last week. “From an industry that produced software, we have become an industry that is manufacturing artificial intelligence.”

Nvidia’s projected sales for the current quarter were encouraging for investors. Many of them are betting that the company’s newest chip, which is faster than its predecessor, will push sales higher.

But the chip’s development has been challenging. During a call with analysts in August, Nvidia said it had to change the production process for the chip because it wasn’t manufacturing enough that worked properly. Mr. Huang assured analysts that wouldn’t cause any delays.

This month, Mr. Huang fulfilled that promise when Nvidia began shipping the first versions of the chip to customers. The premium version, which the company calls the GB200 NVL72, is a computer server with 72 GPUs that weighs 3,000 pounds and has a water-cooling system to prevent it from overheating.

“We’re all betting that there’s a major inflection point coming in A.I. computing, and this is that product,” said David Readerman, a portfolio manager at Endurance Capital Partners, an Nvidia investor. “But it’s a complex product that requires new data center design, installation and performance.”

In addition to those challenges, Nvidia has faced government inquiries into its sales practices. The Justice Department, the European Union, Britain and China have all been looking into its business.

The company has said it will provide any information that regulators need.

In the last quarter, Nvidia continued to dominate the market for A.I. data centers. The company reported that data center sales rose 112 percent to $30.8 billion, with a majority of that coming from the sales of GPUs.

During the call with analysts, Mr. Huang said the company was committed to updating its GPU line annually and was on track to deliver a successor to Blackwell in 2026. He said that was important to increase the performance of its chips for customers, so that they could more efficiently create A.I. systems.

He also said the company, which makes many of its chips and A.I. systems in Taiwan, would support and comply with whatever action the incoming Trump administration took as it considered imposing tariffs on products made overseas.

“We expect this growth, this modernization and this creation of a new industry to go on for several years,” he said.

Explore Our Coverage of Artificial Intelligence

News and Analysis

• Elon Musk has amended a lawsuit he brought against OpenAI, making new antitrust claims against the company and adding defendants, including Microsoft and the venture capitalist Reid Hoffman.
• Meta will allow U.S. government agencies and contractors working on national security to use its A.I. models for military purposes. The shift in policy is intended to promote “responsible and ethical” innovations, the company said.
• A national security memorandum signed by President Biden detailed how the Pentagon, the intelligence agencies and other national security institutions should use and protect A.I. technology.

The Age of A.I.

• A radio station in Poland fired its on-air talent and brought in A.I.-generated presenters. An outcry over a purported chat with a Nobel laureate quickly ended that experiment.
• Meta unveiled a set of A.I. tools for automatically generating videos, instantly editing them and synchronizing them with A.I.-generated sound effects, ambient noise and background music.
• Despite — or, perhaps, because of — the rise in artificially made images, photography is suddenly in the spotlight, in galleries in New York and beyond.