M4 16GB vs RX 7900 XTX

For running local LLMs · 2 reports across 2 models

Side A

Vendor: apple
VRAM: 16GB
Memory: Unified

Side B

Vendor: amd
VRAM: 24GB
Memory: Discrete

Tokens per second by model

Model	M4 16GB	RX 7900 XTX
Gemma 4up to 31B	23.0n=1	—
Qwen3.6up to 35B	—	—n=1

More comparisons