Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
78.7 tok/s
TTFT
2461 ms
Safe context
26K
Memory
9.7 GB / 16.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | B | Runs well | 78.7 tok/s | 3580 ms | 46K |
| Chat | C | Runs well | 78.7 tok/s | 1342 ms | 14K |
| Coding | C | Runs well | 78.7 tok/s | 2461 ms | 26K |
| RAG | B | Runs well | 78.7 tok/s | 4475 ms | 46K |
| Reasoning | C | Runs well | 78.7 tok/s | 2908 ms | 26K |
How Qwen3.5 9B (9B params) fits at each quantization level on Tesla P100 16GB (16.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 3.5 GB | Low | D34 |
Q3_K_S | 3 | 4.4 GB | Low | D35 |
NVFP4 | 4 | 5.0 GB | Medium | D36 |
Q4_K_M | 4 | 5.5 GB | Medium | D37 |
Q5_K_M | 5 | 6.5 GB | High | D38 |
Q6_K | 6 | 7.4 GB | High | D39 |
Q8_0Best for your GPU | 8 | 9.6 GB | Very High | C42 |
F16 | 16 | 18.5 GB | Maximum | F0 |
huggingface-cli download hf-unsloth--qwen3-5-9b-gguf