can it run
3.2 GB needed of 8 GB usable — headroom for context
8GB · Turing TU106 · model file 2.6GB · 4B
computed roofline: 448 GB/s × 0.48–0.72 efficiency window / 2.6GB (Q4_K_M) — bands, never points. Small models (under ~4.5B) run BELOW these bands — kernel-launch overhead keeps them from saturating bandwidth (measured against community medians, 09-06).
→ Tesla M60 (16GB, used) — cheapest catalog machine that runs qwen3 4B fully ($35)
→ RTX 3090 (24GB, used) — upgrade path machine
→ every model the NVIDIA RTX 2070 can run — full list
→explore the full catalog91 machines indexed · live prices · what each one can run