can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA GB300 Grace Blackwell (288GB HBM + 480GB) run Granite 4.2 30B?

Granite · Apache 2.0 · estimated for a 288.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Granite 4.2 30B at FP16 · 58.6 GB · 78–117 tok/s on a NVIDIA GB300 Grace Blackwell (288GB HBM + 480GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K10.9 GB211–316 tok/sRuns great
Q3_K_M14.1 GB189–283 tok/sRuns great
Q4_K_M17.7 GB169–254 tok/sRuns great
Q5_K_M20.8 GB156–233 tok/sRuns great
Q6_K24.0 GB143–215 tok/sRuns great
Q8_031.1 GB122–184 tok/sRuns great
FP1658.6 GB78–117 tok/sRuns great

Ran Granite 4.2 30B on a NVIDIA GB300 Grace Blackwell (288GB HBM + 480GB)? Add your real numbers →