Lawrence Livermore NL Tuolumne

Compute

4.6K H100e

Chips

4,608

First operational

Nov 18, 2024

Lawrence Livermore NL Tuolumne is a large GPU cluster operated by US Department of Energy, located in the United States, in the public sector. It came online around November 2024. At that moment it ranked number 35 among known clusters.

It is built on 4,608 AMD Instinct MI300A. That amounts to about 4.6K H100e of AI compute. The cluster draws roughly 5.66 MW. Its hardware is estimated at $189M.

Epoch AI rates the certainty of this record as confirmed.

Full record

Owner: US Department of Energy
Users: Lawrence Livermore NL, Scientific Research
Sector: Public
Status: Existing
Country: United States
Location: Lawrence Livermore National Laboratory 7000 East Ave, Livermore, CA 94550
Primary chip: AMD Instinct MI300A
Primary chip count: 4,608
Total AI chips: 4,608
H100 equivalents: 4.6K
8-bit throughput: 9×10¹⁸ OP/s
Power capacity: 5.66 MW
Hardware cost: $189M
First operational: Nov 18, 2024
Rank when it came online: 35
Certainty: Confirmed