local intelligence

can it run

Can the AMD Radeon Instinct MI50 (32 GB) run gpt-oss 120B MoE?

NO

65.3 GB vs 32 GB usable — does not fit

32GB · GCN5.1 · model file 65GB · 120B MoE

~29–47 tok/s (active-expert estimate)

computed roofline: 1024 GB/s × 0.28–0.45 efficiency window / 9.8GB (Q4_K_M) — bands, never points.

MoE: file is 65GB (that decides fit) but only ~18B of experts activate per token (~9.8GB streamed) — band below is the active-expert estimate, optimistic; community MoE rows run several× the naive bandwidth math

Run gpt-oss 120B MoE on instead

NIMO AI Mini PC (128GB) — cheapest catalog machine that runs gpt-oss 120B MoE fully ($1999)

RTX 3090 (24GB, used) — upgrade path machine

every model the AMD Radeon Instinct MI50 (32 GB) can run — full list

explore the full catalog84 machines indexed · live prices · what each one can run
2026-09-17 · ← all models on the AMD Radeon Instinct MI50 (32 GB) · fit = weights + 4k KV vs usable memory