Some of Nvidia's largest customers have been told that prices on servers containing its AI chips will rise by more than 15% in many cases, according to Bloomberg News, as soaring memory chip costs squeeze the AI hardware supply chain.
What's going up, and for whom
The increases apply to systems shipped starting early next year and will hit servers built around Nvidia's flagship Vera Rubin and Grace Blackwell chips, with the exact size of the hike depending on chip generation and memory configuration. Server builders that manufacture systems under contract for major data center operators, including Microsoft, Alphabet's Google, and Oracle, have reportedly already notified their customers of the coming increases.
Why: a memory chip crunch
The price pressure traces back to the DRAM market, where three companies, Samsung Electronics, SK Hynix and Micron Technology, dominate production and, according to reporting, still haven't caught up with surging AI-driven demand. Analysts have projected conventional DRAM contract prices climbing as much as 58% to 63% quarter-over-quarter, a spike some in the industry have taken to calling "RAMageddon." Gaming GPU prices have also risen, by as much as 39% in some cases, as the same memory shortage ripples into consumer hardware.
Even with the added costs being passed on to customers, Nvidia has maintained a gross margin near 75%. Major buyers like Amazon, Microsoft, Google and Meta are all developing in-house AI chips, but remain heavily dependent on Nvidia purchases in the meantime.
