Nvidia to Raise AI Server Prices Over 15% for Major Customers

Nvidia Plans Significant Price Increases on AI Systems

Nvidia intends to raise prices on servers equipped with its artificial intelligence processors, marking a notable shift for customers who have already been navigating tight supply and high costs in the AI infrastructure market.

The chipmaker will increase prices by more than 15% in many cases for systems containing its Vera Rubin and Grace Blackwell AI chips, according to a report from Bloomberg News. The exact percentage will vary based on chip generation and memory configurations. These higher prices will apply to systems shipped beginning next year.

Why it matters

The price hikes signal that Nvidia believes demand for its AI accelerators remains strong enough to absorb higher costs, even as competition intensifies from AMD, custom chips from cloud providers, and emerging startups. For enterprises building AI infrastructure, these increases compound already substantial capital expenditures and may accelerate interest in alternative chip architectures or more efficient deployment strategies. The move also highlights how memory chip costs—a critical component often overlooked in discussions of GPU pricing—are creating margin pressure throughout the AI supply chain.

Rising Memory Costs Drive Decision

Nvidia has been contending with escalating costs for memory chips, which are essential components in both its graphics processing units and complete server systems. High-bandwidth memory, in particular, has become a bottleneck in AI system production, with suppliers struggling to meet surging demand from multiple chip manufacturers.

The company's pricing adjustment appears designed to pass these increased component costs through to customers rather than absorbing them and compressing margins. This approach reflects Nvidia's dominant market position in AI accelerators, where alternatives remain limited for many workloads.

Impact on Cloud Providers and Enterprises

The price increases will primarily affect Nvidia's largest customers, which include major cloud service providers like Microsoft, Amazon Web Services, and Google Cloud, as well as large enterprises deploying private AI infrastructure. These organizations have been ordering AI servers in massive quantities to support generative AI services and internal machine learning workloads.

For customers, the higher prices may influence procurement strategies, potentially encouraging longer-term purchase commitments to lock in current pricing or prompting renewed evaluation of competing chip offerings. The timing—with increases taking effect on 2027 shipments—gives customers several months to adjust budgets and deployment plans.

Details of Nvidia's pricing strategy were first reported by Bloomberg News on Saturday.