Money and markets
Nvidia's $96 billion quarter added $442 billion to its value in a day
The chipmaker behind much of today's AI posted record sales on August 26, investors piled in the next day, and fellow chipmaker Broadcom learned a week later that great results don't always get a cheer.
If you have used ChatGPT, Gemini, Claude or almost any other AI assistant, the answer was quite possibly worked out on a chip designed by Nvidia. Its latest results, filed with US regulators on August 26, 2026, were enormous.
What Nvidia actually reported
Nvidia's financial quarter ended on July 26. Over those three months, the company says it took in $96.2 billion in revenue. Revenue is simply the money a company brings in from sales, before any costs are taken off. That figure was up 106% on the same quarter a year earlier. In plain terms, sales more than doubled in a year.
Most of it came from data centers: huge buildings packed with computers that train and run AI models. Nvidia's data-center business alone brought in $89.0 billion. Its gross margin was 75%. Gross margin is the share of each sales dollar left over after paying the direct cost of making the product, so Nvidia kept about 75 cents of every dollar before other expenses.
Companies also give guidance, which is their own forecast for the next quarter. Nvidia said it expects about $108.0 billion in revenue next time. Chief executive Jensen Huang summed up his view in the press release in four words: "Now, compute is revenue." His point is that raw computing power has become something customers will pay for on a vast scale.
A $442 billion day
Market value (or market capitalization) is different from revenue. It is the share price multiplied by the number of shares: a running estimate of what investors think the whole company is worth, changing every trading day.
Nvidia's shares closed at $230.86 on August 26, before the results came out, according to Yahoo Finance's market blog. The next day, August 27, the shares jumped 8.7% and Nvidia added $442 billion to its market value, Bloomberg reported. That was the second-biggest one-day gain by any company's shares in history, behind only a $450 billion jump by Microsoft less than a month earlier. It is a jump in market value, not cash in Nvidia's bank account: investors decided, in a single day, that the company was worth far more than they had thought.
Great results don't always get a cheer
Another AI chip supplier reported strong numbers a week later and got a much cooler reception.
On September 3, Broadcom said its revenue from AI chips rose 221% to $16.7 billion. Yet its shares went roughly flat, because its forecast for the next quarter, $34.8 billion, came in a fraction below the roughly $35 billion Wall Street analysts had expected.
The lesson for newcomers: share prices react to what investors expected, not only to how fast a company is growing.
How to read AI money headlines
- Check the date. A share price is only true for the day it describes.
- Revenue or value? Revenue is money earned. Market value is investors' opinion, and it can swing by hundreds of billions overnight.
- Result or forecast? The $96.2 billion has happened. The $108 billion is a prediction.
- Who said it? An official filing beats a headline, and a headline beats a search snippet.
What we don't know yet
- Nvidia's value today. One search snippet put it at about $5.56 trillion on October 2. That is unconfirmed; check a live quote.
- Who buys the data-center chips. A split between the largest cloud companies and other customers is circulating, but it was not visible in the release we checked.
- Whether Nvidia hits its forecast. We will find out when it next reports.
Sources
- Nvidia press release filed with the SEC (Form 8-K), Aug 26, 2026
- Bloomberg: Nvidia adds $442 billion in second-biggest ever stock surge, Aug 27, 2026
- CNBC: Nvidia adds more than $400 billion in value after blowout earnings, Aug 27, 2026
- Yahoo Finance market live blog, Aug 26, 2026
- Motley Fool on Broadcom's results, Sep 3, 2026
Keep reading
New to AI? Start with which assistant to use, how to stop AI training on your chats, which ChatGPT plan to pay for or how to write a good prompt.