Announced and under-construction GPU clusters
22 clusters. 0 are running or partly running. Installed GPUs on record add up to 10,224, with a further 424,336 planned across 14 clusters. The largest is Nscale Texas Microsoft cluster (104,000 GPUs planned).
Announced counts are intentions. This list keeps them apart from installed GPUs so you can see what is contracted, who it is for and when it is due.
Last verified
22 clusters
Full sortable table| Cluster | Owner / operator | GPU model | GPUs installed | GPUs planned | Location | Status |
|---|---|---|---|---|---|---|
| CoreWeave Barcelona H200 supercomputer | CoreWeave | NVIDIA H200 | 10,224 | — | Barcelona, Spain | announced |
| Nscale Texas Microsoft cluster | Nscale | NVIDIA GB300 | — | 104,000 | Barstow (Cedarvale site), Ward County, Texas, United States | announced |
| Argonne Solstice | Argonne National Laboratory (DOE) | NVIDIA Blackwell | — | 100,000 | Lemont, Illinois, United States | announced |
| Nscale Narvik (Stargate Norway) | Nscale | NVIDIA GB300 | — | 52,000 | Narvik, Norway | under construction |
| Together AI–Hypertec Blackwell cluster | Together AI and Hypertec Cloud | NVIDIA GB200 NVL72 | — | 36,000 | — | announced |
| Yotta Greater Noida D2 Blackwell deployment | Yotta | NVIDIA Blackwell B300 Ultra | — | 30,000 | Greater Noida, India | announced |
| Nscale Loughton AI campus | Nscale | NVIDIA GB300 | — | 23,040 | Loughton, Essex, United Kingdom | announced |
| HUMAIN GB300 AI factory (phase 1) | HUMAIN (Public Investment Fund) | NVIDIA GB300 | — | 18,000 | Saudi Arabia | announced |
| Mistral AI Bruyères-le-Châtel cluster | Mistral AI | NVIDIA GB300 | — | 13,800 | Bruyères-le-Châtel (Eclairion campus), Île-de-France, France | announced |
| Argonne Equinox | Argonne National Laboratory (DOE) | NVIDIA Blackwell | — | 10,000 | Lemont, Illinois, United States | announced |
| Foxconn Taiwan AI Factory (Big Innovation Company) | Big Innovation Company (Foxconn) | NVIDIA Blackwell Ultra (GB300 NVL72) | — | 10,000 | Southern Taiwan | announced |
| KISTI National Supercomputer No. 6 | KISTI (Korea Institute of Science and Technology Information) | NVIDIA GH200 | — | 8,496 | — | under construction |
| Stargate UK (Nscale for OpenAI) | Nscale (OpenAI offtake) | — | — | 8,000 | Cobalt Park, North Tyneside, United Kingdom (one of several UK sites) | announced |
| GMI Cloud Taiwan AI Factory | GMI Cloud | NVIDIA Blackwell | — | 7,000 | Taoyuan, Taiwan | announced |
| Lux (ORNL) | Oak Ridge National Laboratory (DOE) | AMD Instinct MI355X, 288 GB HBM3E | — | 4,000 | Oak Ridge, Tennessee, United States | under construction |
| Alice Recoque | Jules Verne consortium (GENCI, CEA, SURF, GRNET) for EuroHPC JU | AMD Instinct MI430X, 432 GB HBM4 | — | — | TGCC, CEA DAM Île-de-France, Bruyères-le-Châtel, France | under construction |
| Doudna (NERSC-10) | NERSC, Lawrence Berkeley National Laboratory (DOE) | NVIDIA Vera Rubin | — | — | Berkeley, California, United States | announced |
| Lambda Kansas City AI Factory | Lambda | NVIDIA Blackwell Ultra | — | — | Kansas City, Missouri, United States | announced |
| Magic G4 on Google Cloud | Google Cloud (for Magic) | NVIDIA H100 | — | — | — | announced |
| Magic G5 on Google Cloud | Google Cloud (for Magic) | NVIDIA Grace Blackwell | — | — | — | announced |
| Meta Hyperion | Meta | — | — | — | Richland Parish, Louisiana, United States | under construction |
| Nebius Kansas City GPU cluster | Nebius | NVIDIA Hopper | — | — | Kansas City, Missouri, United States | announced |