Horizon Compute Baobab Phase 1

Compute

1K H100e

Chips

1,024

First operational

Feb 15, 2024

Horizon Compute Baobab Phase 1 is a large GPU cluster operated by Horizon Compute, located in the United States, in the private sector. It came online around February 2024. At that moment it ranked number 50 among known clusters.

It is built on 1,024 NVIDIA H100 SXM5 80GB. That amounts to about 1K H100e of AI compute. The cluster draws roughly 1.46 MW. Its hardware is estimated at $45M.

Epoch AI rates the certainty of this record as likely. Judging by their current offerings (March 2025), they have a 2k H100 cluster and a 1k H100 cluster, so I'm interpreting this as them saying that they had two 1k H100 clusters in Q1 2024

Full record

Owner: Horizon Compute
Users: Cloud
Sector: Private
Status: Existing
Country: United States
Primary chip: NVIDIA H100 SXM5 80GB
Primary chip count: 1,024
Total AI chips: 1,024
H100 equivalents: 1K
8-bit throughput: 2×10¹⁸ OP/s
Power capacity: 1.46 MW
Hardware cost: $45M
First operational: Feb 15, 2024
Rank when it came online: 50
Certainty: Likely