H100 80GB vs M5 Max 64GB

For running local LLMs · 5 reports across 2 models

Side A

Vendor: nvidia
VRAM: 80GB
Memory: Discrete

Side B

Vendor: apple
VRAM: 64GB
Memory: Unified

Tokens per second by model

Model	H100 80GB	M5 Max 64GB
Qwen3.6up to 35B	45.0n=1	63.0n=3
Gemma 4up to 31B	—	97.0n=1

More comparisons