A30 vs H100 SXM 80GB
The cheapest A30 ($0.15) costs 70% less per hour than the cheapest H100 SXM 80GB ($0.50).
Per unit of performance, the H100 SXM 80GB is 45% cheaper: $0.17 versus $0.30 per A100-equivalent hour.
The H100 SXM 80GB has 80 GB of memory against 24 GB on the A30.
| Measure | NVIDIA A30 24GB | NVIDIA H100 80GB SXM5 |
|---|---|---|
| Lowest per GPU-hour | $0.15 | $0.50 |
| Median per GPU-hour | $0.64 | $3.51 |
| On-demand from | $0.22 | $0.58 |
| Spot from | $0.15 | $0.50 |
| Reserved from | $0.23 | $1.41 |
| Lowest per A100-equivalent hour | $0.30 | $0.17 |
| Live offers | 32 | 317 |
| Providers | 17 | 83 |
| Memory | 24 GB HBM2 | 80 GB HBM3 |
| Memory bandwidth | 933 GB/s | 3,350 GB/s |
| FP16 / BF16 | 165 TFLOPS | 989 TFLOPS |
| FP8 | — | 1,979 TFLOPS |
| Performance vs A100 80GB | 0.5× | 3× |
| Architecture | Ampere | Hopper |
| Power | 165 W | 700 W |