DallasTexas · US-TX-1
160cards in catalogA dedicated NVIDIA H200 SXM with 141 GB HBM3e, 24 vCPU, 192 GB system RAM, and 3.84 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× H200 SXM | 141 GB | $640 | $2,290 | 0% |
| 2× H200 SXM | 282 GB | $1,216 | $4,351 | 5% |
| 4× H200 SXM | 564 GB | $2,355 | $8,427 | 8% |
| 8× H200 SXM | 1128 GB | $4,608 | $16,488 | 10% |
Regional inventory
Ready-to-run images
Workload fit
Consider this 141 GB configuration when an 80 GB CUDA allocation leaves too little room for weights, runtime state or a useful batch.
Your catalog configuration pairs 141 GB GPU memory with 192 GB host RAM and 3.84 TB local NVMe per card.
Additional capacity is useful only if memory is the binding constraint. Measure whether it avoids sharding or increases sustainable concurrency for your workload.
Use $640 for a 7-day pilot, or $2,290 for a 30-day allocation after validating the workload. Options are additional.