can it run
Comfortably. 748GB of memory against a 4GB working set leaves headroom for context, the OS, and anything else you're running.
| model file | 1.4GB (GGUF Q4_K_M) |
| minimum memory | 4GB working set |
| this machine | 748GB coherent (252GB HBM3e + 496GB LPDDR5x) |
| params | 1.7B |
| license | apache-2.0 |
| price | $94900 |
easy run: ollama pull qwen3:1.7b
runs with: ollama · llama.cpp · koboldcpp · lm studio
no alternatives — every machine in the catalog can run this one.