Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
349.7 tok/s
TTFT
653 ms
Safe context
91K
Memory
50.7 GB / 288.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 354.6 tok/s | 794 ms | 168K |
| Chat | C | Runs well | 354.6 tok/s | 350 ms | 47K |
| Coding | C | Runs well | 349.7 tok/s | 653 ms | 91K |
| RAG | C | Runs well | 354.6 tok/s | 993 ms | 168K |
| Reasoning | C | Runs well | 354.6 tok/s | 645 ms | 91K |
How Qwen3.5 27B (27B params) fits at each quantization level on AMD Instinct MI350X 288GB (288.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 10.5 GB | Low | D30 |
Q3_K_S | 3 | 13.2 GB | Low | D31 |
NVFP4 | 4 |
huggingface-cli download hf-unsloth--qwen3-5-27b-gguf15.1 GB |
| Medium |
| D31 |
Q4_K_M | 4 | 16.5 GB | Medium | D31 |
Q5_K_M | 5 | 19.4 GB | High | D31 |
Q6_K | 6 | 22.1 GB | High | D31 |
Q8_0 | 8 | 28.9 GB | Very High | D32 |
F16Best for your GPU | 16 | 55.4 GB | Maximum | D34 |