gpu4.pics
data via LocalScore
NVIDIA RTX PRO 6000 Blackwell Server Edition
95 GB VRAM NVIDIA
running Llama 3.2 1B Instruct· Q4_K_M· 1.5B
106
tokens / sec
2,962 · Excellent #24 of 768
Generation
106tok/s
higher is better
Prompt / prefill
18,081tok/s
higher is better
Time to first token
74 ms
lower is better
Composite score
2,962
Model size
1.5B
Runtime
gpu4.pics/r/ls-nvidia-rtx-pro-6000-blackwell-server-edition-llama-32-1b-instruct-q4-k-m

All benchmarks on this rig

ModelQuantGenPromptTTFTPowerPerf/W$/MtokScore
Llama 3.2 1B Instruct1.5BQ4_K_M10618,08174 ms2,962

How it stacks up — Llama 3.2 1B Instruct

NVIDIA RTX PRO 6000 Blackwell Server Edition
106
NVIDIA L40S
355
NVIDIA RTX 3090 Ti
354
NVIDIA H100
335
NVIDIA RTX 3090
330
NVIDIA RTX A6000
315
NVIDIA A100 80GB
308