Key Points:
- Nvidia reported $81.6 billion in quarterly sales and expects to reach $91 billion in the next quarter.
- The company announced a huge $80 billion stock repurchase program and increased its quarterly dividend to 25 cents.
- Chief Executive Officer Jensen Huang emphasized a shift toward physical artificial intelligence and selling chips to new businesses.
- Strict export controls and local resistance pushed Nvidia’s data center revenue in China to zero.
Nvidia wants investors to know it can do much more than just sell chips to giant data centers. The company used its latest quarterly report to highlight new efforts to diversify its growing business. While massive tech companies drive most of the current sales, Nvidia executives predict that other businesses and government agencies will soon buy more computing products. These new customers want to support their own artificial intelligence projects and need the hardware to make it happen.
Chief Executive Officer Jensen Huang shared his vision for the future during a recent call with financial analysts. He pointed to physical artificial intelligence as the next huge opportunity for the company. This new category includes smart robots and self-driving vehicles that interact with the real world. Huang confidently told listeners that Nvidia has everything covered for this upcoming technological shift.
Despite these big promises, investors showed a surprising lack of enthusiasm. Nvidia easily beat Wall Street estimates in its latest earnings report and outlook. The company even offered major financial rewards to its shareholders. Executives announced an $80 billion stock repurchase plan and boosted the quarterly dividend from just one penny to 25 cents a share. Even with all this good news, the stock price barely moved when markets opened in New York on Thursday.
The actual financial numbers paint a picture of strong growth. For the three months ending on April 26, Nvidia sales jumped 85% to reach $81.6 billion. This result easily topped the $79.2 billion average estimate from analysts. Profit also climbed to $1.87 a share, beating the $1.77 projection. The company kept 75% of its revenue as gross margin after deducting all production costs.
Looking ahead, Nvidia expects sales to hit about $91 billion for the three months ending in July. This forecast beats the average analyst estimate of $87 billion. However, some highly optimistic analysts had projected numbers as high as $96 billion. This slight miss on the highest market expectations might explain why some investors felt slightly disappointed with the report.
Competition is finally heating up in the artificial intelligence computing market. Nvidia currently rules as the top seller of artificial intelligence accelerators. But other companies across Silicon Valley want a piece of the action. Advanced Micro Devices offers rival processors to eager buyers. Broadcom and Google attack the market with their own custom technology.
Furthermore, the biggest buyers of Nvidia chips now spend billions to develop their own internal components. Smaller chip upstarts are finding success in this booming market as well. Cerebras Systems, a company that builds a unique product based on large silicon chips, held the largest initial public offering of the year just last week.
Even with new rivals entering the race, Nvidia holds a very strong position. Wall Street experts predict that Nvidia will account for more than a third of all sales across the entire semiconductor sector this year. Huang described the current artificial intelligence factory build-out as the largest infrastructure expansion in human history. He noted that this expansion continues to accelerate at an extraordinary speed.
Data center spending remains the biggest cash engine for Nvidia. The massive tech companies, known as hyperscalers, plan to spend a combined $725 billion on artificial intelligence hardware this year. The Nvidia data center unit generated $75.2 billion in revenue last quarter. Within that division, networking equipment delivered $14.8 billion in sales. To help investors track all this money, Nvidia will start using a new reporting framework. The company will separate sales to hyperscalers from a new group called ACIE, which stands for artificial intelligence clouds, industrial, and enterprise customers.
Nvidia does not just sell accelerators anymore. The Santa Clara company now offers central processing units, networking gear, software, and complete computer systems. Engineers also build chips tailored specifically for the inference stage of artificial intelligence. This stage occurs when computer models complete training and begin handling real-world user inputs. Executives expect to bring in $20 billion in central processing unit revenue this year. Hitting that target would make Nvidia the world’s largest supplier of those specific chips. Overall, estimates show Nvidia is on course to record more than $370 billion in total revenue this year. The company regularly reports more sales in a single quarter than its next three largest rivals produce combined.
International politics still create major roadblocks for the chipmaker. Huang recently returned from a trip to China with President Donald Trump. China usually represents the largest overall market for semiconductors. However, strict United States export rules restrict the sale of advanced accelerators on national security grounds. While the Trump administration now allows Nvidia to sell older products to Chinese customers, the government in Beijing resists the offer. Chinese leaders want to build local suppliers instead. Because of this standoff, Nvidia currently makes zero data center revenue in China, missing out on a market that executives say could generate $50 billion a year.
