By Nathan Benaich (Air Street Capital)Reading the inventory
A GPU cluster connects GPU-equipped servers so workloads can use multiple machines together. A GPU fleet is an inventory that can span several clusters or locations; it does not necessarily form one interconnected system.
The index includes both individual clusters and aggregate fleet entries, as identified in the row labels and sources. Compare the same scope and deployment status. GPU generation, memory, networking, and workload also affect performance, so twice as many GPUs does not automatically mean twice as much useful compute.
Citation estimates use inputs through September 1, 2026; fleet counts have an October 1, 2026 cutoff. The research-topic sample covers January 1 to June 1, 2025. See each answer for its period and limitations.
All data sourcesBenaich, Nathan. “What is a GPU cluster, and how is it different from a fleet?” State of AI Report Compute Index. Web page updated 2026-10-11; data periods as specified above.