Servers equipped with Nvidia AI chips are projected to cost approximately 15 percent more, according to Bloomberg. This price hike is attributed to a memory shortage, which is impacting systems using Vera Rubin and Grace Blackwell chips. The shortage is primarily driven by rising DRAM costs from Samsung, SK Hynix, and Micron, with the price increases expected to affect shipments early next year. Contract manufacturers assembling servers for Microsoft, Google, and Oracle have already informed their clients of the anticipated cost rise. Nvidia has not made any official comments on the matter.
The financial burden of these price increases is likely to fall on major cloud providers such as Amazon, Microsoft, Google, and Meta, as well as AI research labs like OpenAI and Anthropic. Despite their efforts to develop in-house chips, these companies still rely heavily on Nvidia's technology. The situation highlights a core tension within the AI industry, where the very firms investing billions in AI infrastructure are also Nvidia's largest customers, supporting the supplier they are attempting to reduce dependence on.
The AI industry continues to require substantial revenue growth to justify these significant investments, as Nvidia carries considerable outstanding liabilities. The current market dynamics underscore the complex interplay between technological advancement and financial sustainability in the AI sector.
Source: thedecoder
