can it run
142.7 GB vs 8 GB usable — does not fit
8GB · Turing TU104 · model file 142GB · 235B-A22B MoE
computed roofline: 448 GB/s × 0.48–0.72 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
→ rtx-3060-12 — the next step up from NVIDIA GeForce RTX 2070 Super
→ every model the NVIDIA GeForce RTX 2070 Super can run — full list
→explore the full catalog84 machines indexed · live prices · what each one can run