Compute per power capacity
Sum H100e / sum MW over the same Existing clusters with both positive values. Calculated capacity takes precedence over disclosed capacity. This is a capacity ratio, not measured efficiency or PUE.
Compute per power capacity
Total 404 · Page 14/17 · Observed 9/25/2026, 3:30:45 AM
| Name | Group | Reported value | Date | Notes | Source |
|---|---|---|---|---|---|
| MIT Supercloud | United States of America | 96.4 H100e/MW | 2020-09-21 | Massachusetts Institute of Technology (MIT) | Source ↗ |
| Immunity Bio Cluster | United States of America | 96.4 H100e/MW | 2020-04-01 | ImmunityBio | Source ↗ |
| Microsoft Azure Immunity Bio | United States of America | 96.4 H100e/MW | 2020-04-01 | Microsoft | Source ↗ |
| Petrobras Atlas | Brazil | 96.4 H100e/MW | 2020-06-01 | Petrobras | Source ↗ |
| Japan Atomic Energy Agency and Quantum and Radiological Science and Technology HPE SGI8600 | Japan | 96.4 H100e/MW | 2020-11-01 | Japan Atomic Energy Agency | Source ↗ |
| Nagoya University Flow Type II (Furo) | Japan | 96.4 H100e/MW | 2020-07-30 | Nagoya University | Source ↗ |
| Jean Zay Supercomputer Phase 2 | France | 96.4 H100e/MW | 2020-06-01 | GENCI | Source ↗ |
| Lawrence Livermore NL Lassen Phase 2 | United States of America | 96.4 H100e/MW | 2020-08-19 | US Department of Energy | Source ↗ |
| Simon Fraser University/Compute Canada Cedar | Canada | 96.4 H100e/MW | 2020-06-01 | Compute Canada | Source ↗ |
| Petrobras Fênix Phase 2 | Brazil | 96.4 H100e/MW | 2020-06-01 | Petrobras | Source ↗ |
| Eni HPC5 | Italy | 96.4 H100e/MW | 2020-02-21 | Eni | Source ↗ |
| Continental DGX Supercomputer | Germany | 96.4 H100e/MW | 2020-07-28 | Continental AG | Source ↗ |
| RPI Supercomputer 2 | United States of America | 96.4 H100e/MW | 2020-06-01 | Rensselaer Polytechnic Institute | Source ↗ |
| Azure OpenAI GPT-3 Cluster | United States of America | 96.4 H100e/MW | 2020-04-15 | Microsoft,OpenAI | Source ↗ |
| MIT LLSC TX-GAIA | United States of America | 94.82 H100e/MW | 2019-09-26 | US Department of Defense | Source ↗ |
| Laboratório Nacional de Computação Científica Santos Dumont | Brazil | 94.82 H100e/MW | 2019-11-01 | Laboratório Nacional de Computação Científica | Source ↗ |
| SENAI CIMATEC Ogbon Cimatec/Petrobras | Brazil | 94.82 H100e/MW | 2019-11-01 | SENAI CIMATEC | Source ↗ |
| CSC Puhti | Finland | 94.82 H100e/MW | 2019-09-03 | CSC – IT Center for Science | Source ↗ |
| University of Tsukuba Cygnus | Japan | 94.82 H100e/MW | 2019-04-30 | University of Tsukuba | Source ↗ |
| Microsoft Azure ND v2 Largest Stated | United States of America | 94.82 H100e/MW | 2019-11-18 | Microsoft | Source ↗ |
| IBM DYEUS | United States of America | 94.82 H100e/MW | 2019-11-18 | IBM | Source ↗ |
| Calcul Québec Béluga | Canada | 94.82 H100e/MW | 2019-04-26 | Compute Canada | Source ↗ |
| Jean Zay Supercomputer Phase 1 | France | 94.82 H100e/MW | 2019-06-01 | GENCI | Source ↗ |
| TotalEnergies Pangea III | France | 94.82 H100e/MW | 2019-06-18 | TotalEnergies | Source ↗ |
| G42 Artemis | United Arab Emirates | 94.82 H100e/MW | 2019-11-01 | G42 | Source ↗ |
Data: Epoch AI · CC BY · View source · Automatically synced; not an exhaustive global sample