Photo by Vadim Bogulov on Unsplash. Source: https://unsplash.com/photos/HgURxw2LMxQ (Unsplash License).

Memory, not the GPU, is now the biggest cost increase inside an AI rack, and Nvidia’s response has been to ship less of it.

  • Memory goes from about 9 percent of a rack’s material cost to about 26 percent between the last two generations.
  • The memory line alone rises from roughly $374,000 to just over $2 million, inside a rack priced near $7.8 million.
  • Nvidia reportedly halved the CPU side memory on its flagship rack, taking each LPDDR5X module from 192GB to 96GB.
  • A new 64GB desktop system ships at $4,999, while the 128GB version rose to $6,950 from a $3,999 launch price.
  • Deloitte expects meaningful new memory fab capacity only in 2029 or 2030.

Read the full breakdown of why the bill of materials flipped and what to lock first.

By Ivan Tarin

Ivan Tarin is a Principal Product Marketing Manager at SUSE, where he owns go-to-market strategy and positioning for a seven-product cloud-native portfolio spanning Kubernetes, virtualization, storage, security, and observability. A former full-stack developer who shipped production code for enterprise and public-sector clients including U.S. national laboratories, Ivan translates complex infrastructure and AI technology into messaging that lands with developers, platform teams, and enterprise buyers. He has presented at KubeCon, SUSECON, and AWS Developer Week, and is currently pursuing an MS in Artificial Intelligence at the University of Colorado Boulder.

Leave a Reply

Your email address will not be published. Required fields are marked *

Get the next one before it is old news

Independent analysis of cloud-native infrastructure, Kubernetes and data center economics. No vendor spin.