RTX 4090 vs RTX 5080
The cheapest RTX 5080 ($0.15) costs 6% less per hour than the cheapest RTX 4090 ($0.16).
Per unit of performance, the RTX 4090 is 17% cheaper: $0.25 versus $0.30 per A100-equivalent hour.
The RTX 4090 has 24 GB of memory against 16 GB on the RTX 5080.
| Measure | NVIDIA GeForce RTX 4090 | NVIDIA GeForce RTX 5080 |
|---|---|---|
| Lowest per GPU-hour | $0.16 | $0.15 |
| Median per GPU-hour | $0.58 | $0.38 |
| On-demand from | $0.19 | $0.15 |
| Spot from | $0.16 | $0.15 |
| Reserved from | $0.18 | $0.25 |
| Lowest per A100-equivalent hour | $0.25 | $0.30 |
| Live offers | 427 | 149 |
| Providers | 91 | 17 |
| Memory | 24 GB GDDR6X | 16 GB GDDR7 |
| Memory bandwidth | 1,008 GB/s | 960 GB/s |
| FP16 / BF16 | 165.2 TFLOPS | 112.5 TFLOPS |
| FP8 | 330.3 TFLOPS | 225.1 TFLOPS |
| Performance vs A100 80GB | 0.63× | 0.5× |
| Architecture | Ada Lovelace | Blackwell |
| Power | 450 W | 360 W |