▲ local intelligence

can it run

Can the NVIDIA RTX A6000 run llava 13B?

FULL

11.1 GB needed of 48 GB usable — headroom for context

48GB · Ampere · model file 8.0GB · 13B

~44–73 tok/s

computed roofline: 768 GB/s × 0.46–0.76 efficiency window / 8.0GB (Q4_K_M) — bands, never points.

Run llava 13B on instead

→ Tesla M60 (16GB, used) — cheapest catalog machine that runs llava 13B fully ($35)

→ RTX 4090 (24GB, used) — upgrade path machine

→ every model the NVIDIA RTX A6000 can run — full list

→explore the full catalog91 machines indexed · live prices · what each one can run
2026-10-01 · ← all models on the NVIDIA RTX A6000 · fit = weights + 4k KV vs usable memory