can it run
Yes, but close to the edge. 384GB against a 363GB working set: it runs, with little room for long context windows or other processes.
weights 290GB + KV cache 0.97GB at 4k context vs ~96GB usable → no. 291.0 GB vs 96 GB usable — does not fit
computed roofline: 1792 GB/s × 0.46–0.76 efficiency window / 290.0GB (Q4_K_M) — bands, never points.
| model file | 290GB (GGUF Q4_K_M) |
| minimum memory | 363GB working set |
| this machine | 384GB VRAM (4× 96GB) |
| params | 480B A35B cloud-class |
| license | apache-2.0 |
| price | $91000 |
easy run: ollama pull qwen3-coder:480b
runs with: ollama · llama.cpp · koboldcpp · lm studio
→ ASUS ExpertCenter Pro ET900N G3 (GB300) — cheapest machine that runs it ($60000)
→ ASUS ExpertCenter Pro ET900N G3 (GB300) — this model on the biggest machine