Nvidia AI Server Prices Set to Rise as Memory Costs Surge

Prices for servers powered by Nvidia artificial intelligence chips are expected to climb sharply as rising memory chip costs put additional pressure on the booming AI infrastructure market, according to a Bloomberg News report.

Some of Nvidia’s largest customers could face server price increases of more than 15%, with the higher costs expected to affect systems delivered in early 2027. The increases will reportedly apply to servers featuring Nvidia’s flagship Vera Rubin and Grace Blackwell processors, although the exact price adjustment will depend on the chip generation and memory configuration.

Server manufacturers supplying major data center operators, including Microsoft, Google and Oracle, have recently informed customers about the planned increases, according to the report.

A key factor behind the higher Nvidia AI server prices is soaring demand for DRAM memory. Samsung Electronics, SK Hynix and Micron Technology dominate global DRAM production and have gained greater pricing power as technology companies accelerate spending on AI data centers. Nvidia’s AI accelerators rely heavily on high-performance memory, making memory supply and pricing increasingly important to overall server costs.

Chip shortages are already affecting other parts of the technology industry. Apple and Qualcomm have recently indicated that supply constraints have forced them to raise prices.

Nvidia, meanwhile, continues to benefit from exceptionally strong demand for its AI processors. The company has a gross margin of about 75% and can charge tens of thousands of dollars for individual chips as manufacturing capacity from Taiwan Semiconductor Manufacturing Co. struggles to keep pace with demand. Nvidia has also reportedly increased prices for gaming-focused graphics cards.

Amazon, Microsoft, Google and Meta are developing proprietary AI chips to reduce their reliance on Nvidia, but the companies still depend heavily on Nvidia hardware for ongoing data center expansion. Their ability to shift toward in-house processors will also depend on securing sufficient memory supplies from Samsung, SK Hynix and Micron.

Higher hardware costs could further complicate the global AI data center boom. Developers are already dealing with project delays, labor shortages, tighter capital markets and community opposition, while rising chip and memory prices threaten to push infrastructure budgets even higher.