can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB) run DeepSeek-R1-Distill 14B?

DeepSeek · Apache 2.0 · estimated for a 144.0 GB machine at 4K context.

Yes — it runs great.

Best fit: DeepSeek-R1-Distill 14B at FP16 · 29.6 GB · 90–135 tok/s on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K5.8 GB220–330 tok/sRuns great
Q3_K_M7.3 GB201–301 tok/sRuns great
Q4_K_M9.0 GB184–276 tok/sRuns great
Q5_K_M10.5 GB171–256 tok/sRuns great
Q6_K12.1 GB159–238 tok/sRuns great
Q8_015.7 GB137–206 tok/sRuns great
FP1629.6 GB90–135 tok/sRuns great

Ran DeepSeek-R1-Distill 14B on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB)? Add your real numbers →