GPU Cloud
H100 (Infiniband)
H100 Configuration with Infiniband (3.2 Tbit/s): 2 Intel Xeon 8480+ / 2TB RAM / 8x3.84 TB NVMe / 8x Nvidia H100, IB 3.2 Tbit/s, 2x100Gbit/s Ethernet
| H100每GPU价格 | 1个月滚动 | 保留6个月 | 保留12个月 | 保留36个月 |
|---|
| 超过512个GPU | €3.30/小时 | €3.15/小时 | €3.05/小时 | €2.90/小时 |
| 128至512个GPU | €3.60/小时 | €3.30/小时 | €3.25/小时 | €3.10/小时 |
| 32 至128个GPU | €3.70/小时 | €3.50/小时 | €3.40/小时 | €3.25/小时 |
| 少于32个GPU | €3.75/小时 | €3.60/小时 | €3.50/小时 | €3.35/小时 |
Prices do not include VAT.
A100 (Infiniband)
A100 80GB with Infiniband 800Gbit/s: 2 Intel Xeon 8468 / 2 TB RAM / 8x3.84 TB NVMe / 8x Nvidia A100 / 800Gbit/s Infiniband
| H100每GPU价格 | 1个月滚动 | 保留6个月 | 保留12个月 | 保留36个月 |
|---|
| 128至512个GPU | €1.98/小时 | €1.60/小时 | €1.58/小时 | €1.30/小时 |
| 32 至128个GPU | €2.03/小时 | €1.70/小时 | €1.65/小时 | €1.36/小时 |
| 少于32个GPU | €2.06/小时 | €1.74/小时 | €1.70/小时 | €1.40/小时 |
Prices do not include VAT.
A100
A100 80GB without Infiniband: 2 Intel Xeon 8468 / 2 TB RAM / 8x3.84 TB NVMe / 8x Nvidia A100, 2x100Gbit/s Ethernet
| H100每GPU价格 | 1个月滚动 | 保留6个月 | 保留12个月 | 保留36个月 |
|---|
| 128至512个GPU | €1.92/小时 | €1.54/小时 | €1.52/小时 | €1.25/小时 |
| 32 至128个GPU | €1.97/小时 | €1.64/小时 | €1.59/小时 | €1.31/小时 |
| 少于32个GPU | €2.00/小时 | €1.70/小时 | €1.64/小时 | €1.35/小时 |
Prices do not include VAT.
L40s
L40s Configuration: 2x Intel Xeon 8468 / 2TB RAM / 4x7.68TB NVMe RI SSD + 2x960SSD OS / 8x Nvidia L40S / 2x10Gbit/s Ethernet
| L40s | 1个月滚动 | 保留6个月 | 保留12个月 | 保留36个月 |
|---|
| 超过512个GPU | €1.64/小时 | €1.50/小时 | €1.41/小时 | €1.28/小时 |
| 128至512个GPU | €1.91/小时 | €1.64/小时 | €1.59/小时 | €1.46/小时 |
| 32 至128个GPU | €2.00/小时 | €1.82/小时 | €1.73/小时 | €1.59/小时 |
| 少于32个GPU | €2.05/小时 | €1.91/小时 | €1.82/小时 | €1.69/小时 |
Prices do not include VAT.
GB200
H200 80GB with Infiniband 3.2Tbit/s
2 Intel Xeon 8480+ / 2TB RAM / 2x 960GB SSD + 8x3.84 TB NVMe / 8x Nvidia H200, IB 3.2 Tbit/s, 2x100Gbit/s Ethernet
| Price per GPU | 1 month Rolling | 6 months Reserved | 12 months Reserved | 36months Reserved |
|---|
| Greater than 512 GPUs | €3.40/hour | €3.25/hour | €3.15/hour | €3.00/hour |
| 128 to 512 GPUs | €3.70/hour | €3.40/hour | €3.35/hour | €3.20/hour |
| 32 to 128 GPUs | €3.80/hour | €3.60/hour | €3.50/hour | €3.35/hour |
| Less than 32 GPUs | €3.85/hour | €3.70/hour | €3.60/hour | €3.45/hour |
H200 GPUs available upon request. Contact us for availability and pricing.
Prices do not include VAT.
GB200
Available upon request. Contact us for availability and pricing.
Everywhere Inference
Billing is determined by the exact compute time you use, measured down to the minute. Benefit
from effortless autoscaling, fast cold starts, and a wide range of GPU options, along with secure
deployments and protected endpoints - all while keeping costs low.
Prices do not include VAT.
For third-party cloud and on-premise deployments please get in touch with our sales team.