llamaperf

H100 80GB vs RTX 5090

For running local LLMs · 13 reports across 3 models

Side A
H100 80GB
Vendor
nvidia
VRAM
80GB
Memory
Discrete
Side B
RTX 5090
Vendor
nvidia
VRAM
32GB
Memory
Discrete

Tokens per second by model

ModelH100 80GBRTX 5090
Qwen3.6up to 35B45.0n=13238.0n=7
Gemma 4up to 31B125.3n=1578.0n=3
Qwen3.5up to 27Bn=1

More comparisons