Parts/Accelerators/AMD Instinct MI325X
Why this part
A memory refresh of MI300X rather than a new compute generation, which is a straightforward read of where the competition actually is. 256GB is more high-bandwidth memory per device than anything NVIDIA shipped in the same window, and for inference that is frequently the only number that decides the deployment.
Systems using this part
- Systems
- 1
- Rmax underneath
- 0.0 GFlop/s
- Units recorded
- 8,192
1 of 1 edges disclose a quantity
| System | Role | Country | Units | System Rmax | Supplied by | Rank |
|---|---|---|---|---|---|---|
| TensorWave Tucson MI325X Cluster 2025-07 | Accelerator | USA | 8,192 | - | AMD | - |
“Units recorded” sums only the edges where a public source states a quantity. It is a floor, not a total, and should never be read as installed base.