A dedicated NVIDIA RTX 5090 with 32 GB GDDR7, 16 vCPU, 128 GB system RAM, and 2 TB local NVMe per card.
Node pricing
The product model allocates each GPU to one account for the complete term. Adding cards to a node applies the catalog volume discount to every card.
| Node | Total VRAM | 7 days | 30 days | Discount |
|---|---|---|---|---|
| 1× RTX 5090 | 32 GB | $124 | $432 | 0% |
| 2× RTX 5090 | 64 GB | $236 | $821 | 5% |
| 4× RTX 5090 | 128 GB | $456 | $1,590 | 8% |
| 8× RTX 5090 | 256 GB | $893 | $3,110 | 10% |
Regional inventory
Ready-to-run images
Workload fit
A 32 GB step between the 24 GB cards and 48 GB configurations for image workflows or local-model experiments needing more headroom.
Your catalog configuration pairs 32 GB GPU memory with 128 GB host RAM and 2 TB local NVMe per card.
Measure whether 32 GB is sufficient before optimizing for generation time. Confirm that your selected image supports the exact GPU and software combination.
Use $124 for a 7-day pilot, or $432 for a 30-day allocation after validating the workload. Options are additional.