Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
224.3 tok/s
TTFT
863 ms
Safe context
41K
Memory
4.7 GB / 12.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 149.5 tok/s | 1884 ms | 82K |
| Chat | C | Runs well | 149.5 tok/s | 706 ms | 21K |
| Coding | C | Runs well | 224.3 tok/s | 863 ms | 41K |
| RAG | C | Runs well | 149.5 tok/s | 2354 ms | 82K |
| Reasoning | C | Runs well | 149.5 tok/s | 1530 ms | 41K |
How EXAONE 3.5 2.4B Instruct (2.4000000953674316B params) fits at each quantization level on Intel Arc B580 12GB (12.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 0.9 GB | Low | D31 |
Q3_K_S | 3 | 1.2 GB | Low | D31 |
NVFP4 | 4 |
| Medium |
| D31 |
Q4_K_M | 4 | 1.5 GB | Medium | D32 |
Q5_K_M | 5 | 1.7 GB | High | D32 |
Q6_K | 6 | 2.0 GB | High | D33 |
Q8_0 | 8 | 2.6 GB | Very High | D34 |
F16Best for your GPU | 16 | 4.9 GB | Maximum | D38 |