# The largest neoclouds, ranked by capacity

Ranked by data-center power. Live (operational) and announced/planned capacity are ranked separately so a plan is never presented as a running site. Power is the comparable axis because most large operators no longer disclose GPU counts. Source: https://registry.deploy.report/largest-neoclouds

## By live (operational) power

| Operator | Region | Live power | GPU fleet (where disclosed) | Accelerator |
|---|---|---|---|---|
| CoreWeave | North America | 1,500 MW | ~250,000 (last disclosed, end-2024) | NVIDIA H100 / H200 / GB200 NVL72 |
| YTL AI Cloud | Asia | 600 MW | NVIDIA GB200 NVL72 (count undisclosed) | - |
| Nebius | Europe | 170 MW | ~60,000 (reported) | NVIDIA H100 / H200 / B200 / GB200 |
| Sify Technologies | Asia | 129 MW | - | NVIDIA (H100 / H200 / GB200; deployed count undisclosed) |
| SK Telecom GPUaaS | Asia | 80 MW | 1,000+ B200 (Haein cluster) | NVIDIA B200 (Petasus) |
| Core42 | Middle East | 60 MW | up to 35,000 GB300-equivalent (US export-approved); 10,000 via Northern Data | NVIDIA H100 / H200 (+ AMD, Cerebras WSE-3) |
| Voltage Park | North America | 60 MW | ~24,000 NVIDIA H100 (~35,000+ combined post-Lightning merger) | NVIDIA H100 (B200 / GB300 added) |
| ePLDT | Asia | 50 MW | - | NVIDIA H100 / H200 |
| Freedom Holding | Asia | 48 MW | - | NVIDIA exascale (Sovereign AI Hub) |
| Ooredoo | Middle East | 26 MW | - | NVIDIA Hopper / H200 |
| GreenNode | Asia | 20 MW | 1,024 H100 | NVIDIA H100 |
| du | Middle East | 13 MW | - | NVIDIA GPUs |

## By announced / planned power

| Operator | Region | Announced power | GPU fleet (where disclosed) | Accelerator |
|---|---|---|---|---|
| Nebius | Europe | 5,000 MW | ~60,000 (reported) | NVIDIA H100 / H200 / B200 / GB200 |
| Crusoe | North America | 4,900 MW | - | NVIDIA GB200 / B200 / H200 + AMD MI300X |
| CoreWeave | North America | 4,200 MW | ~250,000 (last disclosed, end-2024) | NVIDIA H100 / H200 / GB200 NVL72 |
| Lambda | North America | 3,000 MW | >1,000,000 target by 2030 | NVIDIA H100 / B200 / GB300 NVL72 |
| TensorWave | North America | 1,000 MW | 8,192 AMD Instinct MI325X (largest AMD cluster in North America) | AMD Instinct MI325X / MI300X / MI355X |
| IREN | North America | 750 MW | ~150,000 target | - |
| Sesterce | Europe | 600 MW | 500,000 GPUs (Grand Est target) | NVIDIA H100 / H200 / B200 |
| HUMAIN | Middle East | 500 MW | 18,000 NVIDIA GB300 (first phase) | NVIDIA GB300 (Grace Blackwell Ultra) |
| Neysa | Asia | 400 MW | 20,000-25,000 accelerators (target) | NVIDIA H100 / L40S |
| Fluidstack | North America | 360 MW | - | NVIDIA H100 / H200 / B200 / GB200 |
| Northern Data | Europe | 250 MW | ~22,000 NVIDIA H100 / H200 (>85% allocated) | NVIDIA H100 / H200 |
| Nscale | Europe | 240 MW | ~200,000 NVIDIA GB300 (Microsoft, phased 2026-27) | NVIDIA GB300 / GB200 / H100 + AMD MI300X |
| Viettel | Asia | 140 MW | 96 H200 + 22 DGX B200 | NVIDIA H200 + DGX B200 (~1.5 exaflops) |
| Sharon AI | Oceania | 132 MW | 40,000 GB300 (target) | NVIDIA H200 + AMD Instinct MI300X |
| Jio Platforms | Asia | 120 MW | >200,000 H100-equivalent at full build (H100-equivalent, not raw count) | NVIDIA GB300 (Blackwell Ultra) |
| du | Middle East | 120 MW | - | NVIDIA GPUs |
| Ooredoo | Middle East | 120 MW | - | NVIDIA Hopper / H200 |
| Freedom Holding | Asia | 100 MW | - | NVIDIA exascale (Sovereign AI Hub) |
| Siam AI | Asia | 100 MW | 1,024 H100 | NVIDIA H100 (to H200 / GB200) |
| Sustainable Metal Cloud | Asia | 84 MW | ~54,000 NVIDIA GB300 (Project Southgate) | NVIDIA H100 (GB300 planned); immersion-cooled |
| NeevCloud | Asia | 80 MW | 60,000 GPUs (target) | NVIDIA H100 / H200 / B200 / B300 |
| WhiteFiber | North America | 76 MW | ~4,000 GPUs (contracted) | NVIDIA H200 / B200 / GB200 |
| ResetData | Oceania | 72 MW | 1,024 H200 (AI-F1) | NVIDIA H200 |
| Mistral Compute | Europe | 44 MW | ~18,000 GPUs | NVIDIA GB300 |
| L&T Vyoma | Asia | 30 MW | 10,000 B300 (Chennai) | NVIDIA H100 / H200 / B200 (B300 deploying) |
| Krutrim | Asia | 20 MW | - | NVIDIA GB200 NVL72 / H100 |
| GMI Cloud | Asia | 16 MW | 7,000 NVIDIA GB300 (Taiwan AI Factory, go-live ~Mar 2026) | NVIDIA GB300 NVL72 |

## GPU fleet disclosed, power not

- Denvr Dataworks: ~1,024 NVIDIA H100 SXM5
- Ishan Technologies: 1,000+ GPUs
- Tata Communications: tens of thousands of NVIDIA H100 (ordered)
- Scaleway: ~2,000+ NVIDIA H100 (Nabuchodonosor)
- Together AI: 36,000 NVIDIA GB200 NVL72 (co-build); access to 100,000+
- Polarise: ~10,000 GPUs (Munich AI Factory)
- Alem.Cloud: 512 H200 (64 HGX servers, ~2 exaflops, TOP500 #86)
- BUZZ HPC: 504-GPU cluster (initial)
- Capitalonline: 20,000+ GPUs (company-wide)
- Highrise AI: 1,000 H100 (Chicago)
- Hot Aisle: 128 MI300X
- Ori: 1,024 H100 (private cloud)
- RUTILEA: 352 H100 (Fukushima)
- KT Cloud: 1,600 MI250
- Ubilink: ~1,024 H100 (SuperPOD, ~45.8 PFLOPS)
- Lyceum: ~8,000 GPUs (target)
- NexGen Cloud: AI Supercloud ~20,000 H100 target (~$1B)
- SenseCore: 54,000+ GPUs (reported)
- Cassava Technologies: ~3,000 GPUs (South Africa)
- Lingjing Cloud: ~5,000 cards (Wuxi Xinwei center)
- GMO GPU Cloud: 768 H200 (TOP500 #37)
- ESDS: 8,208 NVIDIA B300 (GPUaaS)
- Yotta Data Services: ~10,000+ live; ordered 16,000 H100/GH200
- Sakura Internet: ~4,000+ (H100 / H200 / Blackwell); ~10,800 planned
- E2E Networks: ~4,000+ (2x1,024 H200 + 1,024 B200 + ~1,000 H100)
- Cyfuture: ~1,184 GPUs (IndiaAI)
- Digi Power X: ~2,304 GPUs (target)

Full definition and roster: https://registry.deploy.report/neoclouds