Some of Nvidia Corp.’s biggest customers have been told that the prices of servers containing its artificial intelligence chips will increase by more than 15% in many cases, driven by soaring memory chip costs. According to Bloomberg News, the price adjustments will affect systems shipped early next year, impacting configurations built around flagship processors including the Vera Rubin and Grace Blackwell chips.
Supply Chain Pressures and Memory Costs
The upcoming price increases depend heavily on the specific generation of Nvidia hardware and the associated memory configurations, according to individuals familiar with the communications who spoke to Bloomberg News. Contract manufacturers responsible for assembling the servers for major data center operators—such as Microsoft Corp., Alphabet Inc.’s Google, and Oracle Corp.—have recently notified their clients of the impending adjustments.
Nvidia representatives did not respond to requests for comment regarding the communications, which have not yet been made public. The developments highlight mounting cost pressures across the semiconductor supply chain. Despite being one of the most profitable firms in the industry, Nvidia has been unable to completely absorb the rising cost of dynamic random access memory, or DRAM, pointing to powerful leverage held by memory component producers.
The Role of Memory Manufacturers
Samsung Electronics Co., SK Hynix Inc., and Micron Technology Inc. control the vast majority of global DRAM production. Although these manufacturers have ramped up factory output, supply has failed to keep pace with runaway global demand for artificial intelligence infrastructure. This persistent shortage has driven up prices for commodity-like memory components, granting manufacturers unprecedented influence within the technology sector.
Nvidia’s accelerator processors serve as the foundation for computers that train and run artificial intelligence software. Their overall effectiveness relies directly on the volume of DRAM they are paired with. While Nvidia continues to command tens of thousands of dollars per chip due to supply limitations at Taiwan Semiconductor Manufacturing Co., rising input costs have forced pricing adjustments further down the supply line.
Other major technology firms, including Apple Inc. and Qualcomm Inc., have also reported being forced to charge higher prices for their own hardware due to ongoing chip shortages. Meanwhile, Nvidia recently implemented price increases for its gaming-oriented personal computer graphics cards, as reported earlier this month by industry news site Tom’s Hardware.
Market Impact and Future Outlook
Major data center operators including Amazon, Microsoft, Google, and Meta continue to pursue proprietary in-house silicon programs, yet they remain heavily dependent on Nvidia hardware for their primary infrastructure build-outs.
The scheduled price increases threaten to add new layers of complexity to ambitious AI data center expansion plans, which have already faced hurdles from project delays, labor shortages, tightening capital markets, and community resistance. Nvidia is scheduled to report its fiscal second-quarter earnings next week, providing investors and industry observers with a vital update on the financial health of the broader artificial intelligence infrastructure market.
