can it run
9.8 GB vs 8 GB usable — partial CPU offload, expect large speed loss
8GB · Turing TU106 · model file 9.0GB · distill 14B
computed roofline: 448 GB/s × 0.48–0.72 efficiency window / 9.0GB (Q4_K_M) — bands, never points.
→ Tesla P100 (16GB, used) — cheapest catalog machine that runs DEEPSEEK-R1 distill 14B fully ($135)
→ RTX 3090 (24GB, used) — upgrade path machine
→ every model the NVIDIA RTX 2070 can run — full list
→explore the full catalog84 machines indexed · live prices · what each one can run