Building 1GW AI Data Center with Nvidia's Vera Rubin Costs $47 Billion, Bernstein Report Shows

According to the latest Bernstein research report, constructing a 1GW AI data center using Nvidia's next-generation Vera Rubin architecture, expected to be deployed at scale in 2027, would require total capital expenditure of approximately $47 billion. A single Vera Rubin NVL72 rack is estimated to cost $9.1 million, higher than the previously cited $8 million benchmark, driven by rising high-bandwidth memory (HBM) prices and increased costs for storage, networking, and power systems. Building a 1GW facility would require roughly 3,557 racks, with equipment costs totaling around $32 billion and additional infrastructure spending of $15 billion for land, buildings, substations, and cooling systems.
Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments