Explore
Side-by-side specs and tok/s estimates for llama-3.1-8b. 13 GPUs in catalog.
| GPU | VRAM | Bandwidth | TDP | Price | tok/s | Confidence |
|---|---|---|---|---|---|---|
| NVIDIA GeForce RTX 5090 | 32 GB | 1792 GB/s | 600 W | — | — | — |
| NVIDIA GeForce RTX 5080 | 16 GB | 1024 GB/s | 400 W | — | — | — |
| NVIDIA GeForce RTX 4070 Ti SUPER | 16 GB | 672 GB/s | 285 W | — | — | — |
| NVIDIA GeForce RTX 4060 Ti 16GB | 16 GB | 288 GB/s | 165 W | — | — | — |
| NVIDIA RTX A6000 | 48 GB | 768 GB/s | 300 W | — | — | — |
| AMD Radeon RX 7900 XTX | 24 GB |
| 960 GB/s |
| 355 W |
| — |
| — |
| — |
| NVIDIA A100 80GB PCIe | 80 GB | 2039 GB/s | 250 W | — | — | — |
| NVIDIA GeForce RTX 4090 | 24 GB | 1008 GB/s | 450 W | — | 163.8 | ESTIMATED |
| NVIDIA GeForce RTX 3090 (Used) | 24 GB | 936 GB/s | 350 W | — | 135.7 | ESTIMATED |
| NVIDIA RTX PRO 6000 Blackwell 96GB | 96 GB | 1792 GB/s | 300 W | — | 246.4 | ESTIMATED |
| NVIDIA RTX PRO 5000 Blackwell 48GB | 48 GB | 896 GB/s | 250 W | — | 123.2 | ESTIMATED |
| NVIDIA GeForce RTX 5070 Ti | 16 GB | 896 GB/s | 300 W | — | 123.2 | ESTIMATED |
| NVIDIA GeForce RTX 4080 SUPER | 16 GB | 736 GB/s | 320 W | — | 114.1 | ESTIMATED |