Written by Jakub Rusinowski · Last updated April 5, 2025
Yes, but it is tight
Yes, but it is tight — Llama 4 Scout 17B at Q8_0 needs about 118.2 GB of the 128 GB usable on Framework Desktop (Ryzen AI Max+ 395, 128 GB), leaving only ~9.8 GB before the runtime starts swapping. Expect ~9.9 tok/s (estimated), with room for about 32,768 tokens of context.
Confidence: medium · Recommended quantization: Q8_0 · Estimated speed: ~9.9 tok/s
| Usable memory for models | 128 GB |
| Memory bandwidth | 256 GB/s |
| Form factor | Mini PC |
| Operating system | Windows or Linux |
| Memory upgradeable | No — soldered |
| Quant | Memory needed | Fits 128 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 220.4 GB | ✗ No | — | — | 218 GB |
| Q8_0 | 118.2 GB | ✓ Yes | 32K | ~9.9 tok/s | 115.8 GB |
| Q6_K | 91.8 GB | ✓ Yes | 128K | ~12.7 tok/s | 89.4 GB |
| Q5_K_M | 79.7 GB | ✓ Yes | 128K | ~14.5 tok/s | 77.3 GB |
| Q4_K_M | 68.2 GB | ✓ Yes | 256K | ~16.7 tok/s | 65.8 GB |
| Q3_K_M | 48.9 GB | ✓ Yes | 256K | ~22.6 tok/s | 46.5 GB |
| Q2_K | 38.2 GB | ✓ Yes | 256K | ~28.1 tok/s | 35.8 GB |
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| Llama 4 Maverick 17B | 243.9 GB | ✗ Too large | — |
| Llama 4 Scout 17B | 68.2 GB | ✓ Fits | ~16.7 tok/s |
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving
Yes, but it is tight — Llama 4 Scout 17B at Q8_0 needs about 118.2 GB of the 128 GB usable on Framework Desktop (Ryzen AI Max+ 395, 128 GB), leaving only ~9.8 GB before the runtime starts swapping. Expect ~9.9 tok/s (estimated), with room for about 32,768 tokens of context.
Q8_0 — it needs about 118.2 GB of the 128 GB available, downloads as roughly 115.8 GB, and runs at an estimated 9.9 tokens/sec with up to 32K of context.
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving