GPU clusters
MTS Grom
Compute
50.45 H100e
Chips
160
First operational
Jun 1, 2021
MTS Grom is a large GPU cluster operated by MTS, located in Russia, in the private sector. It came online around June 2021. At that moment it ranked number 91 among known clusters.
It is built on 160 NVIDIA A100. That amounts to about 50.45 H100e of AI compute. The cluster draws roughly 0.14 MW. Its hardware is estimated at $4M.
Epoch AI rates the certainty of this record as confirmed.
Full record
- Owner
- MTS
- Users
- MTS, Cloud
- Sector
- Private
- Status
- Existing
- Country
- Russia
- Location
- Russia
- Primary chip
- NVIDIA A100
- Primary chip count
- 160
- Total AI chips
- 160
- H100 equivalents
- 50.45
- 8-bit throughput
- 10×10¹⁶ OP/s
- Power capacity
- 0.14 MW
- Hardware cost
- $4M
- First operational
- Jun 1, 2021
- Rank when it came online
- 91
- Certainty
- Confirmed
SourceEpoch AI, 'GPU Clusters'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/gpu-clusters. Licensed under CC BY 4.0.
← All gpu clusters