This page answers unsloth/Qwen3.5-397B-A17B-GGUF q4_k_m quantization queries with explicit calculations from our model requirement dataset and compatibility speed table.
Q4_K_M mapped to the same VRAM envelope as Q4 in current dataset.
Throughput data below uses available compatibility measurements/estimates and is sorted by tokens per second for this model.
Need general guidance? Review full methodology.
| GPU | VRAM | Quantization | Speed | Compatibility |
|---|---|---|---|---|
| AMD Instinct MI300X | 192GB | Q4 | 100 tok/s | View full compatibility |
| NVIDIA H200 SXM 141GB | 141GB | Q4 | 81 tok/s | View full compatibility |
| NVIDIA H100 SXM5 80GB | 80GB | Q4 | 62 tok/s | View full compatibility |
| AMD Instinct MI250X | 128GB | Q4 | 58 tok/s | View full compatibility |
| NVIDIA H100 PCIe 80GB | 80GB | Q4 | 39 tok/s | View full compatibility |
| RTX 5090 | 32GB | Q4 | 35 tok/s | View full compatibility |
| NVIDIA A100 80GB SXM4 | 80GB | Q4 | 33 tok/s | View full compatibility |
| AMD Instinct MI210 | 64GB | Q4 | 30 tok/s | View full compatibility |
| NVIDIA A100 40GB PCIe | 40GB | Q4 | 28 tok/s | View full compatibility |
| NVIDIA RTX 6000 Ada | 48GB | Q4 | 22 tok/s | View full compatibility |
| RTX 4090 | 24GB | Q4 | 20 tok/s | View full compatibility |
| RTX 5080 | 16GB | Q4 | 20 tok/s | View full compatibility |