Nvidia shed roughly $200 billion of market value last week after OpenAI disclosed a $50 billion revenue run rate, $18 billion below the figure that had been circulating among investors. The chipmaker's shares fell 4% on the news, a modest percentage move that carries outsized weight for a company valued near $5 trillion.
The number that moved the market
OpenAI, which remains private and does not file quarterly results, chose to reveal a run-rate figure that fell short of the $68 billion whisper number. Because the startup is a bellwether for data-center demand, the miss was read as a signal that the pace of AI infrastructure spending could slow. Nvidia, which supplies the majority of the accelerators behind large-model training and inference, absorbed the brunt of the repricing.
Why the chipmaker is still central
Nvidia's newest platform, Vera Rubin, combines Rubin GPUs, Vera CPUs and large pools of high-bandwidth memory into full compute systems. The company said revenue more than doubled year over year in the fiscal second quarter ended July 26, and management has guided for 70% growth in fiscal 2028. That trajectory would be unusual for a business of this scale, but it remains the baseline the market is pricing.
Valuation and the buying question
At less than 14 times forward one-year earnings, the stock is not expensive by historical standards. The 4% decline does not on its own create a dramatic entry point, but it may offer a slightly better level for investors who were already planning to add exposure. The real uncertainty is whether OpenAI's slower-than-expected revenue ramp portends a broader deceleration in accelerator demand, or simply reflects the lumpiness of a private company's disclosure choices.