can it run
Yes, but close to the edge. 8GB against a 7GB working set: it runs, with little room for long context windows or other processes.
weights 4.9GB + KV cache 0.5GB at 4k context vs ~6GB usable → tight. 5.4 of 6 GB usable — keep context modest
computed roofline: 68 GB/s × 0.25–0.65 efficiency window / 4.9GB (Q4_K_M) — bands, never points.
| model file | 4.9GB (GGUF Q4_K_M) |
| minimum memory | 7GB working set |
| this machine | 8GB unified |
| params | 8B |
| license | llama3.1 (gated on HF) |
| price | $499 |
easy run: ollama pull llama3.1:8b
runs with: ollama · llama.cpp · koboldcpp · lm studio
→ Tesla P100 (16GB, used) — cheapest machine that runs it ($135)
→ every model the Jetson Orin Nano Super dev kit can run — full list