CoreWeave H200s

Compute

42K H100e

Chips

42,000

First operational

Aug 28, 2024

CoreWeave H200s is a large GPU cluster operated by CoreWeave, located in the United States, in the private sector. It came online around August 2024. At that moment it ranked number 1 among known clusters.

It is built on 42,000 NVIDIA H200 SXM. That amounts to about 42K H100e of AI compute. The cluster draws roughly 59.93 MW. Its hardware is estimated at $1.5B.

Epoch AI rates the certainty of this record as likely.

Full record

Owner: CoreWeave
Users: CoreWeave
Sector: Private
Status: Existing
Country: United States
Primary chip: NVIDIA H200 SXM
Primary chip count: 42,000
Total AI chips: 42,000
H100 equivalents: 42K
8-bit throughput: 8.3×10¹⁹ OP/s
Power capacity: 59.93 MW
Hardware cost: $1.5B
First operational: Aug 28, 2024
Rank when it came online: 1
Certainty: Likely