can-i-run-this-llm

Can I run this LLM?

Can a AMD Instinct MI300X (192GB) run Ornith 1.5 35B-A3B (MoE)?

Ornith · MIT · estimated for a 192.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Ornith 1.5 35B-A3B (MoE) at FP16 · 71.1 GB · 234–350 tok/s on a AMD Instinct MI300X (192GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K13.1 GB330–495 tok/sRuns great
Q3_K_M16.7 GB322–482 tok/sRuns great
Q4_K_M21.9 GB311–466 tok/sRuns great
Q5_K_M25.5 GB303–455 tok/sRuns great
Q6_K30.5 GB293–440 tok/sRuns great
Q8_037.8 GB281–421 tok/sRuns great
FP1671.1 GB234–350 tok/sRuns great

Ran Ornith 1.5 35B-A3B (MoE) on a AMD Instinct MI300X (192GB)? Add your real numbers →