A100 40GB vs RTX 3090

For running local LLMs · 3 reports across 2 models

Side A

Vendor: nvidia
VRAM: 40GB
Memory: Discrete

Side B

Vendor: nvidia
VRAM: 24GB
Memory: Discrete

Tokens per second by model

Model	A100 40GB	RTX 3090
Qwen3.6up to 35B	—	66.0n=2
Qwen2.5	—	28.0n=1

More comparisons