SANTA CLARA, CALIFORNIA / RankWire.AI / – Nvidia intends to increase prices by more than 15% for numerous AI server configurations expected to ship in early 2027. These hikes apply to setups utilizing the company’s Vera Rubin and Grace Blackwell platforms, with the magnitude of each increase differing depending on chip generation and memory configuration. Nvidia has not officially declared a blanket price hike across all server products, though server manufacturers have already shared the revised pricing details with key data center clients.
Major data center operators such as Microsoft, Google, and Oracle have been informed of the updated pricing by server manufacturers, who have relayed this information directly. These companies purchase substantial quantities of AI infrastructure for cloud and data center operations. The rising costs of memory have become a significant factor influencing the reported increases in server pricing. AI deployments heavily depend on large amounts of high-bandwidth memory, along with GPUs, CPUs, networking hardware, and storage. During 2026, demand from AI applications has kept several memory categories in tight supply.
TrendForce predicts that contract prices for conventional DRAM will increase between 13% and 18% in the third quarter of 2026, while NAND Flash contract prices are expected to rise by 10% to 15% over the same period. The firm noted that server DRAM remains undersupplied because suppliers are prioritizing capacity for AI-related projects. Although long-term supply agreements have limited some price hikes, overall server memory expenses continue to be elevated.
Memory Price Surge Continues Amid Persistent AI Server Demand
In May, Nvidia announced that Vera Rubin had reached full production with server manufacturers and supply-chain partners worldwide. The company indicated that products integrating Rubin would be available during the latter half of 2026. Vera Rubin combines the Vera CPU, Rubin GPU, NVLink 6, ConnectX-9, BlueField-4, and Spectrum-6 networking technologies. It succeeds Grace Blackwell as Nvidia’s latest rack-scale architecture designed for large AI workloads. System manufacturers are developing Rubin-based solutions for cloud providers, AI research labs, and hyperscale data centers.
Meanwhile, Grace Blackwell remains a key component of Nvidia’s current data center offerings. The GB200 NVL72 configuration connects 36 Grace CPUs with 72 Blackwell GPUs within a single liquid-cooled rack. Designed to function as a unified NVLink computing domain, it caters to large-scale AI models. The reported increases in server prices differ depending on specific configurations, with memory capacity and chip generation being crucial factors. Therefore, the 15% increase does not apply uniformly across all Nvidia server models.
Expansion of Vera Rubin Production Shapes Nvidia’s Server Ecosystem
Memory suppliers have shifted more production capacity towards server and high-performance products as AI demand continues to absorb available supply. According to TrendForce, this shift has resulted in reduced availability of certain PC and consumer DRAM products. Server shipments remained robust throughout 2026, as buyers continued building inventories of data center memory. The analyst expects supply tightness for server DRAM to persist into 2027, with demand growth surpassing new supply, creating a high-cost environment that influences Nvidia’s announced pricing adjustments.
Nvidia enters this pricing cycle with record-breaking data center revenue. In its fiscal first quarter ending April 26, the company reported total sales of $81.6 billion, with data center revenue reaching an all-time high of $75.2 billion—a 92% increase year-over-year. Nvidia forecasted second-quarter revenue of approximately $91 billion, plus or minus 2%, during its May earnings report. The company is scheduled to release its fiscal second-quarter results on Aug. 26.