llamaperf

M5 Max 64GB vs RTX 4090

For running local LLMs · 6 reports across 2 models

Side A
M5 Max 64GB
Vendor
apple
VRAM
64GB
Memory
Unified
Side B
RTX 4090
Vendor
nvidia
VRAM
24GB
Memory
Discrete

Tokens per second by model

ModelM5 Max 64GBRTX 4090
Qwen3.6up to 35B63.0n=280.0n=2
Gemma 4up to 31B97.0n=1149.6n=1

More comparisons