can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB) run MiMo V2.6 Distill Qwen 9B?

MiMo · MIT · estimated for a 144.0 GB machine at 4K context.

Yes — it runs great.

Best fit: MiMo V2.6 Distill Qwen 9B at FP16 · 17.9 GB · 130–194 tok/s on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K3.6 GB265–398 tok/sRuns great
Q3_K_M4.5 GB250–375 tok/sRuns great
Q4_K_M5.8 GB229–343 tok/sRuns great
Q5_K_M6.9 GB214–322 tok/sRuns great
Q6_K7.8 GB203–305 tok/sRuns great
Q8_09.6 GB185–278 tok/sRuns great
FP1617.9 GB130–194 tok/sRuns great

Ran MiMo V2.6 Distill Qwen 9B on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB)? Add your real numbers →