AI GPU Platforms
Rack-scale and pod-scale configurations for AI training and inference. Aggregate PFLOPS, power draw, and interconnect — side by side.
Filters
Available Systems
| Compare | ||||||
|---|---|---|---|---|---|---|
NVIDIA Vera Rubin NVL72 NVL· rack | 72× Rubin SXM Rubin | 1260.0 | 288.0 | 9.4 | — | |
NVIDIA DGX B300 DGX· rack | 8× B300 SXM 288GB Blackwell Ultra | 40.0 | 20.0 | 0.7 | 14.5 kW | |
NVIDIA DGX GB300 NVL72 NVL· rack | 72× B300 SXM 288GB Blackwell Ultra | 360.0 | 180.0 | 6.0 | — | |
NVIDIA DGX SuperPOD GB200 SuperPOD· pod | 2304× B200 SXM 180GB Blackwell | 10368.0 | 5184.0 | 172.8 | 3840.0 kW | |
NVIDIA GB200 NVL32 NVL· rack | 32× B200 SXM 180GB Blackwell | 144.0 | 72.0 | 2.4 | 53.3 kW | |
NVIDIA GB200 NVL36 NVL | 36× B200 SXM 180GB Blackwell | 162.0 | 81.0 | 2.7 | 60.0 kW | |
NVIDIA GB200 NVL576 NVL | 576× B200 SXM 180GB Blackwell | 2592.0 | 1296.0 | 43.2 | 960.0 kW | |
NVIDIA GB200 NVL72 NVL· rack | 72× B200 SXM 180GB Blackwell | 324.0 | 162.0 | 5.4 | 120.0 kW | |
NVIDIA HGX B200 8-GPU HGX· baseboard | 8× B200 SXM 180GB Blackwell | 36.0 | 18.0 | 0.6 | 10.2 kW | |
NVIDIA DGX H200 DGX· rack | 8× H200 SXM 141GB Hopper | 15.8 | 7.9 | 0.5 | 8.5 kW | |
NVIDIA HGX B100 8-GPU HGX· baseboard | 8× B100 SXM 192GB Blackwell | 28.0 | 14.0 | 0.5 | 5.6 kW | |
NVIDIA HGX H200 HGX· baseboard | 8× H200 SXM 141GB Hopper | 15.8 | 7.9 | 0.5 | 8.0 kW | |
NVIDIA DGX H100 DGX· rack | 8× H100 SXM5 80GB Hopper | 15.8 | 7.9 | 0.5 | 8.5 kW | |
NVIDIA HGX H100 4-GPU HGX· baseboard | 4× H100 SXM5 80GB Hopper | 7.9 | 4.0 | 0.3 | 4.2 kW | |
NVIDIA HGX H100 8-GPU HGX· baseboard | 8× H100 SXM5 80GB Hopper | 15.8 | 7.9 | 0.5 | 8.0 kW | |
NVIDIA DGX A100 320GB DGX· rack | 8× A100 PCIe 40GB Ampere | — | 2.5 | 0.2 | 6.0 kW | |
NVIDIA DGX A100 640GB DGX· rack | 8× A100 SXM4 80GB Ampere | — | 2.5 | 0.2 | 6.5 kW | |
NVIDIA HGX A100 4-GPU HGX· baseboard | 4× A100 SXM4 80GB Ampere | — | 1.2 | 0.1 | 3.2 kW | |
NVIDIA HGX A100 8-GPU HGX· baseboard | 8× A100 SXM4 80GB Ampere | — | 2.5 | 0.2 | 6.0 kW |
per page
1/1
Rent These Systems
Cloud providers offering managed instances with these GPU systems.
| Provider | Instance | GPU | Price/hr | |
|---|---|---|---|---|
RunPod | H100 SXM | H100 SXM5 80GB | $2.69 | View → |
DataCrunch | 1H100.80S.30V | H100 SXM5 80GB | $3.25 | View → |
RunPod | A100 SXM | A100 SXM4 80GB | $1.39 | View → |
DataCrunch | 1A100.22V | A100 SXM4 80GB | $1.79 | View → |
Beyond.pl | DGX B200 | B200 SXM 180GB | Contact | View → |
Vast.ai | — | B200 SXM 180GB | $4.63 | View → |
RunPod | B200 | B200 SXM 180GB | $5.89 | View → |
DataCrunch | 1B200.30V | B200 SXM 180GB | $6.11 | View → |
Vast.ai | — | H200 SXM 141GB | $2.76 | View → |
RunPod | H200 SXM | H200 SXM 141GB | $3.59 | View → |
DataCrunch | 1H200.141S.44V | H200 SXM 141GB | $4.00 | View → |
RunPod | B300 MIG 34GB | B300 SXM 288GB | $0.50 | View → |
Vast.ai | — | B300 SXM 288GB | $6.70 | View → |
DataCrunch | 1B300.30V | B300 SXM 288GB | $7.50 | View → |