Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
111.0 tok/s
TTFT
1744 ms
Safe context
22K
Memory
23.8 GB / 32.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | B | Runs well | 111.0 tok/s | 2537 ms | 43K |
| Chat | B | Runs well | 111.0 tok/s | 951 ms | 11K |
| Coding | B | Runs well | 111.0 tok/s | 1744 ms | 22K |
| RAG | B | Runs well | 111.0 tok/s | 3172 ms | 43K |
| Reasoning | B | Runs well | 111.0 tok/s | 2062 ms | 22K |
How Qwen 3 30B A3B (30.5B params) fits at each quantization level on AMD Instinct MI100 32GB (32.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 11.9 GB | Low | D37 |
Q3_K_S | 3 | 14.9 GB | Low | D39 |
NVFP4 | 4 |
ollama run qwen-3-30b-a3bhuggingface-cli download qwen-3-30b-a3b17.1 GB |
| Medium |
| C41 |
Q4_K_M | 4 | 18.6 GB | Medium | C42 |
Q5_K_M | 5 | 22.0 GB | High | C44 |
Q6_KBest for your GPU | 6 | 25.0 GB | High | C44 |
Q8_0 | 8 | 32.6 GB | Very High | F0 |
F16 | 16 | 62.5 GB | Maximum | F0 |