can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA RTX 6000 Ada (48GB) run Llama 3.3 70B Instruct?

Llama · Llama 3.3 Community · estimated for a 48.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Llama 3.3 70B Instruct at Q3_K_M · 34.3 GB · 19–28 tok/s on a NVIDIA RTX 6000 Ada (48GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K26.4 GB24–36 tok/sRuns great
Q3_K_M34.3 GB19–28 tok/sRuns great
Q4_K_M42.5 GB15–23 tok/sWon't fit
Q5_K_M50.0 GB13–20 tok/sWon't fit
Q6_K57.9 GB12–17 tok/sWon't fit
Q8_075.0 GB9–14 tok/sWon't fit
FP16141.1 GB5–7 tok/sWon't fit

Ran Llama 3.3 70B Instruct on a NVIDIA RTX 6000 Ada (48GB)? Add your real numbers →