Nvidia AI Server Prices and Rising Memory Costs
Nvidia AI server prices could rise by more than 15% as soaring memory costs squeeze the next generation of artificial-intelligence infrastructure. This Short explains what the reported increases could mean for Vera Rubin and Grace Blackwell systems, hyperscalers, enterprise AI buyers, and the global data-center boom.
Reuters reports that some of Nvidia’s largest customers have been warned about higher prices on systems shipping early next year. Contract server manufacturers working for major data-center operators—including Microsoft, Alphabet’s Google, and Oracle—have reportedly begun informing customers. The exact increase may vary by chip generation and memory configuration.
A mid-teens jump in server prices can have an enormous impact when companies are building clusters worth billions of dollars. Higher costs may affect cloud-computing prices, AI model training, inference economics, startup access to compute, data-center financing, and demand for custom chips.
Reuters said it could not independently verify the initial Bloomberg report, and Nvidia did not immediately respond outside regular business hours.
Source: Reuters — August 22, 2026
Reporting: Disha Mishra
Editing: Franklin Paul
What to watch
The reported increase may vary by system and memory configuration. It should not be treated as a confirmed price list for every buyer. The original reporting’s verification limits matter when considering possible effects on cloud prices, training budgets, and access to computing capacity.
Related reading: our discussion of AI infrastructure financing.
Watch and listen
Watch the YouTube Short above or listen to the full episode on Spotify.
