Autor: Jakub Rusinowski · Ostatnia aktualizacja: 15 czerwca 2026
Yes — the Llama 3.2 Family family runs on the Intel Arc B580.
| VRAM | 12 GB |
| Memory Bandwidth | 456 GB/s |
| Llama 3.2 11B Vision Instruct | Q4_K_M · 7.8 GB · ~58 tok/s (est.) |
| Llama 3.2 3B Instruct | Q4_K_M · 2.2 GB · ~207 tok/s (est.) |
| Llama 3.2 1B Instruct | Q4_K_M · 0.8 GB · ~400 tok/s (est.) |
Yes — the Llama 3.2 Family family runs on the Intel Arc B580.
Llama 3.2 11B Vision Instruct at Q4_K_M quantization (7.8 GB), estimated ~58 tokens/sec.
← Can I Run It? | Llama 3.2 Family Model Page | Intel Arc B580 GPU Page | Check Your Hardware