Nvidia’s primary customers are bracing for a significant increase in costs, with reports indicating that servers equipped with the company’s advanced AI chips may see price hikes exceeding 15 percent. This adjustment is largely driven by the surging cost of memory chips, which are essential components for high-performance AI infrastructure. According to insiders, these new price structures will apply to systems shipping early next year, including those integrated with the highly anticipated Grace Blackwell and Vera Rubin architectures.

The price increases are expected to vary based on the specific generation of hardware and memory configuration, affecting the major contractors who supply server units to tech giants like Microsoft, Google, and Oracle. As Nvidia continues to serve as the backbone for global AI development, these rising costs highlight the escalating expense of maintaining massive data center operations. With Nvidia’s second-quarter earnings report scheduled for August 26, the industry is watching closely to see how these supply chain pressures and pricing shifts will impact the broader AI ecosystem.