local intelligence

can it run

Can the Radeon PRO W7900 48GB run llama3.1 8B?

FULL

5.4 GB needed of 48 GB usable — headroom for context

48GB · RDNA3 · model file 4.9GB · 8B

~88–130 tok/s

computed roofline: 864 GB/s × 0.50–0.75 efficiency window / 4.9GB (Q4_K_M) — bands, never points.

Run llama3.1 8B on instead

Tesla M60 (16GB, used) — cheapest catalog machine that runs llama3.1 8B fully ($35)

rtx-pro-6000 — the next step up from Radeon PRO W7900 48GB

every model the Radeon PRO W7900 48GB can run — full list

explore the full catalog90 machines indexed · live prices · what each one can run
2026-09-24 · ← all models on the Radeon PRO W7900 48GB · fit = weights + 4k KV vs usable memory