can it run
Yes, but close to the edge. 192GB against a 178GB working set: it runs, with little room for long context windows or other processes.
weights 142GB + KV cache 0.73GB at 4k context vs ~190GB usable → full. 142.7 GB needed of 190 GB usable — headroom for context
computed roofline: 256 GB/s × 0.60–0.85 efficiency window / 13.3GB (Q4_K_M) — bands, never points.
MoE: file is 142GB (that decides fit) but only ~22B of experts activate per token (~13.3GB streamed) — band below is the active-expert estimate, optimistic; community MoE rows run several× the naive bandwidth math
| model file | 142GB (GGUF Q4_K_M) |
| minimum memory | 178GB working set |
| this machine | 192GB LPDDR5X-8533 unified (160GB GPU-assignable) |
| params | 235B-A22B MoE |
| license | apache-2.0 |
| price | $None |
easy run: ollama pull qwen3:235b-a22b
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
HotHardware (announce) ↗ · Amazon ↗