Kovara
Loading your page…
You can continue using the navigation while this page loads.
You can continue using the navigation while this page loads.
H100 NVL 94GB has 33% less memory than H200 SXM (94 GB vs 141 GB). Its memory bandwidth is 19% less (3.9 vs 4.8 TB/s). In the cloud, the median on-demand H100 NVL 94GB price is $8.66 per GPU-hour against $4.47 for the H200 SXM. Per gigabyte of GPU memory, that is $0.092 vs $0.032 per hour.
Manufacturer figures as recorded by Kovara. Higher memory and bandwidth are highlighted; neither is a benchmark of your workload.
| H100 NVL 94GB | H200 SXM | |
|---|---|---|
| Manufacturer | NVIDIA | NVIDIA |
| Architecture | Hopper | Hopper |
| Memory | 94 GB HBM3 | 141 GB HBM3E |
| Memory bandwidth | 3.9 TB/s | 4.8 TB/s |
| Power (TDP) | 400 W | 700 W |
| Form factor | PCIe | SXM |
Public offers Kovara tracks, normalised to US dollars per GPU-hour. Prices use on-demand terms where available; spot and reserved offers are on each GPU’s price pages. Lower is highlighted.
| H100 NVL 94GB | H200 SXM | |
|---|---|---|
| Lowest price | $2.92 | $2.60 |
| Median price | $8.66 | $4.47 |
| Highest price | $10.12 | $10.60 |
| Median per GB of memory | $0.092 | $0.032 |
| Public offers (all terms) | 60 | 41 |
| Providers | 3 | 14 |
Lowest tracked price per provider. Open a provider to see every offer, region and term.
Right now the cheapest on-demand H100 NVL 94GB Kovara tracks is $2.92 per GPU-hour and the cheapest H200 SXM is $2.60. Medians are $8.66 and $4.47. The cheaper GPU per hour is not always cheaper per job: a faster or larger GPU can finish sooner or need fewer units.
The H200 SXM has more memory: 141 GB against 94 GB. Memory decides which models and batch sizes fit on one GPU.
No. This page compares manufacturer specifications and public prices. Real performance depends on the model, precision, software and interconnect. Kova can help you estimate cost for a specific workload.