can it run
170.9 GB vs 48 GB usable — does not fit
48GB · Ampere · model file 168.9GB · 552B+196B Engram MoE
computed roofline: 768 GB/s × 0.46–0.76 efficiency window / 25.3GB (Q4_K_M) — bands, never points.
MoE: file is 168.9GB (that decides fit) but only ~82.8B of experts activate per token (~25.3GB streamed) — band below is the active-expert estimate, optimistic; community MoE rows run several× the naive bandwidth math
→ RTX 4090 (24GB, used) — upgrade path machine
→ every model the NVIDIA RTX A6000 can run — full list
→explore the full catalog84 machines indexed · live prices · what each one can run