llamaperf

H100 80GB vs M2 Ultra 192GB

For running local LLMs · 4 reports across 4 models

Side A
H100 80GB
Vendor
nvidia
VRAM
80GB
Memory
Discrete
Side B
M2 Ultra 192GB
Vendor
apple
VRAM
192GB
Memory
Unified

Tokens per second by model

ModelH100 80GBM2 Ultra 192GB
DeepSeek V328.0n=1
Gemma 4up to 31B125.3n=1
Qwen3.5up to 122Bn=1
Qwen3.6up to 35B45.0n=1

More comparisons