GPU comparison
H200 vs H100
H200 and H100 share the same Hopper compute (1.98 PF FP8), so the question is memory. H200 has 141 GB at 4.8 TB/s against 80 GB at 3.35 TB/s: if your model, batch or context does not fit in 80 GB, the H200 saves you GPUs and is faster on memory-bound inference.
- Updated Nov 4, 2026
- Specs: vendor datasheets, dense Tensor figures
- Prices: live CoreGPU catalog
| Spec | NVIDIA H200 SXM | NVIDIA H100 SXM |
|---|---|---|
| Architecture | Hopper | Hopper |
| GPU memory | 141 GB HBM3e | 80 GB HBM3 |
| Memory bandwidth | 4.8 TB/s | 3.35 TB/s |
| FP8 dense | 1.98 PF | 1.98 PF |
| FP4 dense | n/a | n/a |
| Interconnect | NVLink 4 · 900 GB/s | NVLink 4 · 900 GB/s |
| Power (TDP) | 700 W | 700 W |
| Price per GPU / month | $2,699 | $1,999 |
| Price per GB of memory | $19.14 | $24.99 |
| Price per FP8 PFLOPS | $1,363 | $1,010 |
| Availability | 34 available | Only 3 left |
Green marks the better value on each line. Prices are per GPU per month at CoreGPU, at least 25% below the market median.
- 70B models at FP8 on a single GPU
- Memory-bound inference and long contexts
- Larger fine-tuning batches
$2,699 /GPU/moConfigure H200
- Models that fit comfortably in 80 GB
- The lowest price for an 8-GPU NVLink node
- Research, CI and CUDA development
$1,999 /GPU/moConfigure H100