can it run
No. 24GB of memory can't hold the 54GB working set this model needs. A smaller quant or a bigger machine is required.
weights 43GB + KV cache 1.25GB at 4k context vs ~24GB usable → no. 44.2 GB vs 24 GB usable — does not fit
computed roofline: 288 GB/s × 0.28–0.50 efficiency window / 43.0GB (Q4_K_M) — bands, never points.
| model file | 43GB (GGUF Q4_K_M) |
| minimum memory | 54GB working set |
| this machine | 24GB GDDR5, 288 GB/s |
| params | 70B |
| license | llama3.1 (gated on HF) |
| price | $55 |
easy run: ollama pull llama3.1:70b
runs with: ollama · llama.cpp · koboldcpp · lm studio
→ NIMO AI Mini PC (128GB) — cheapest machine that runs it ($1999)
→ every model the Tesla M40 (24GB, used) can run — full list
→ tesla-p40 — the next step up from this machine's GPU