can-i-run-this-llm

Can I run this LLM?

Can a AMD Instinct MI300X (192GB) run Gemma 4 12B?

Gemma · Apache 2.0 · estimated for a 192.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Gemma 4 12B at FP16 · 23.8 GB · 112–168 tok/s on a AMD Instinct MI300X (192GB).

Every quantization on this machine

QuantSizeSpeedFit
Q3_K_M5.7 GB234–350 tok/sRuns great
Q4_K_M7.1 GB215–323 tok/sRuns great
Q5_K_M8.4 GB201–301 tok/sRuns great
Q6_K9.8 GB188–281 tok/sRuns great
Q8_012.7 GB165–247 tok/sRuns great
FP1623.8 GB112–168 tok/sRuns great

Ran Gemma 4 12B on a AMD Instinct MI300X (192GB)? Add your real numbers →