Written by Jakub Rusinowski · Last updated November 20, 2024
Yes — comfortably
Yes, comfortably — SmolLM2 1.7B Instruct at Q8_0 needs about 4.2 GB of the 12.8 GB usable on 16 GB system RAM, leaving ~8.6 GB spare and running at ~24.2 tok/s (estimated), with room for about 8,192 tokens of context.
Confidence: high · Recommended quantization: Q8_0 · Estimated speed: ~24.2 tok/s
| Usable memory for models | 12.8 GB |
| Memory bandwidth | 90 GB/s |
| Quant | Memory needed | Fits 12.8 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 5.8 GB | ✓ Yes | 8K | ~15.4 tok/s | 3.4 GB |
| Q8_0 | 4.2 GB | ✓ Yes | 8K | ~24.2 tok/s | 1.8 GB |
| Q6_K | 3.8 GB | ✓ Yes | 8K | ~28.5 tok/s | 1.4 GB |
| Q5_K_M | 3.6 GB | ✓ Yes | 8K | ~31 tok/s | 1.2 GB |
| Q4_K_M | 3.4 GB | ✓ Yes | 8K | ~33.8 tok/s | 1 GB |
| Q3_K_M | 3.1 GB | ✓ Yes | 8K | ~39.9 tok/s | 0.7 GB |
| Q2_K | 3 GB | ✓ Yes | 8K | ~44.3 tok/s | 0.6 GB |
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| SmolLM2 1.7B Instruct | 3.4 GB | ✓ Fits | ~33.8 tok/s |
| SmolLM2 360M Instruct | 1.4 GB | ✓ Fits | ~125.1 tok/s |
llama.cpp (CPU build) or Ollama — both run without a GPU
Yes, comfortably — SmolLM2 1.7B Instruct at Q8_0 needs about 4.2 GB of the 12.8 GB usable on 16 GB system RAM, leaving ~8.6 GB spare and running at ~24.2 tok/s (estimated), with room for about 8,192 tokens of context.
Q8_0 — it needs about 4.2 GB of the 12.8 GB available, downloads as roughly 1.8 GB, and runs at an estimated 24.2 tokens/sec with up to 8K of context.
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
llama.cpp (CPU build) or Ollama — both run without a GPU