gpu4.pics
data via LocalScore
NVIDIA A100 80GB
80 GB VRAM NVIDIA 400 W TDP
running Llama 3.2 1B Instruct· Q4_K_M· 1.5B
308
tokens / sec
3,892 · Excellent #8 of 768
Generation
308tok/s
higher is better
Prompt / prefill
15,370tok/s
higher is better
Time to first token
80 ms
lower is better
Composite score
3,892
Model size
1.5B
Runtime
gpu4.pics/r/ls-nvidia-a100-80gb-llama-32-1b-instruct-q4-k-m

All benchmarks on this rig

ModelQuantGenPromptTTFTPowerPerf/W$/MtokScore
Llama 3.2 1B Instruct1.5BQ4_K_M30815,37080 ms3,892

How it stacks up — Llama 3.2 1B Instruct

NVIDIA A100 80GB
308
NVIDIA L40S
355
NVIDIA RTX 3090 Ti
354
NVIDIA H100
335
NVIDIA RTX 3090
330
NVIDIA RTX A6000
315