Nvidia has reported another record quarter, with revenue doubling from a year earlier to 96.2 billion dollars as demand for the chips that power artificial intelligence showed no sign of easing. The figure, for the three months to late July, comfortably beat the roughly 92.2 billion dollars analysts had expected.
The engine of that growth remains its data centre business, which generated about 89 billion dollars, up 117 percent on a year earlier and ahead of forecasts. Adjusted earnings came in at 2.22 dollars per share, above the 2.10 dollars Wall Street had penciled in.
Nvidia’s guidance was, if anything, more striking than the results. It told investors to expect revenue of around 108 billion dollars in the current quarter, above the 104.2 billion consensus, signalling that the buildout of AI computing power is still accelerating.
“AI has reached its inflection point,” chief executive Jensen Huang said, adding that compute had become revenue and that demand was accelerating. He pointed to a broadening base of customers, with multiple frontier research labs and open-model developers now driving orders.
There were notes of caution. Nvidia said gross margins, currently around 75 percent, would ease to between 71 and 72 percent by the end of the fiscal year as it ramps up its next-generation Vera Rubin platform.
Despite beating estimates on every major measure, Nvidia’s shares slipped about 1.8 percent in after-hours trading, a reminder that for a company valued so richly, merely meeting extraordinary expectations is sometimes not enough. Still, the quarter cements its position at the heart of the AI economy.
