Parts/Accelerators/NVIDIA A100 (in DGX A100)
Accelerators · Ampere
NVIDIA A100 (in DGX A100)
NVIDIA A100 Tensor Core GPU as fitted to DGX A100 nodes, recorded without a memory variant.
Why this part
Recorded because Meta's Research SuperCluster is described only as DGX A100 systems and the sources do not say whether the GPUs carry 40 GB or 80 GB, so the existing memory-specific A100 parts are not used. Per-chip figures are not on the pages we fetched.
Systems using this part
- Systems
- 13
- Rmax underneath
- 95.65 PFlop/s
- Units recorded
- 6,849
5 of 13 edges disclose a quantity
| System | Role | Country | Units | System Rmax | Supplied by | Rank |
|---|---|---|---|---|---|---|
| Sejong 2023 | Accelerator | South Korea | - | 32.97 PFlop/s | NVIDIA | #22 → |
| University of Nottingham Malaysia DGX A100 2022-02 | Accelerator | Malaysia | - | - | NVIDIA | - |
| KT DGX SuperPOD 2022 | Accelerator | South Korea | - | 10.38 PFlop/s | NVIDIA | #58 → |
| Cambridge-1 2021-07-07 | Accelerator | United Kingdom | 640 | 9.68 PFlop/s | NVIDIA | #41 → |
| BioHive-1 2021-06 | Accelerator | USA | - | 4.90 PFlop/s | NVIDIA | #84 → |
| HiPerGator AI 2021 | Accelerator | USA | - | 17.49 PFlop/s | NVIDIA | #106 → |
| MTS GROM 2021 | Accelerator | Russia | - | 2.26 PFlop/s | NVIDIA | #240 → |
| IARA 2021 | Accelerator | Brazil | 25 | 3.66 PFlop/s | NVIDIA | #234 → |
| NVIDIA DGX SuperPOD (A100) 2020 | Accelerator | USA | - | 2.36 PFlop/s | NVIDIA | #170 → |
| Meta Research SuperCluster (RSC) | Accelerator | USA | 6,080 | - | NVIDIA | - |
| Christofari Neo | Accelerator | Russia | - | 11.95 PFlop/s | NVIDIA | #43 → |
| Serbia National AI Platform | Accelerator | Serbia | 32 | - | NVIDIA | - |
| Technion Zeus | Accelerator | Israel | 72 | - | NVIDIA | - |
“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.