H100 SXM 80GB vs V100 32GB
The cheapest V100 32GB ($0.017) costs 97% less per hour than the cheapest H100 SXM 80GB ($0.50).
Per unit of performance, the V100 32GB is 76% cheaper: $0.04 versus $0.17 per A100-equivalent hour.
The H100 SXM 80GB has 80 GB of memory against 32 GB on the V100 32GB.
| Measure | NVIDIA H100 80GB SXM5 | NVIDIA Tesla V100 32GB |
|---|---|---|
| Lowest per GPU-hour | $0.50 | $0.017 |
| Median per GPU-hour | $3.51 | $2.32 |
| On-demand from | $0.58 | $0.017 |
| Spot from | $0.50 | $0.51 |
| Reserved from | $1.41 | $0.19 |
| Lowest per A100-equivalent hour | $0.17 | $0.04 |
| Live offers | 317 | 193 |
| Providers | 83 | 43 |
| Memory | 80 GB HBM3 | 32 GB HBM2 |
| Memory bandwidth | 3,350 GB/s | 900 GB/s |
| FP16 / BF16 | 989 TFLOPS | 125 TFLOPS |
| FP8 | 1,979 TFLOPS | — |
| Performance vs A100 80GB | 3× | 0.42× |
| Architecture | Hopper | Volta |
| Power | 700 W | 300 W |