AMD Instinct MI250X~105 tok/s (Q4)

Qwen/Qwen2.5-Math-72B-Instruct speed on AMD Instinct MI250X

Quantization-specific throughput and VRAM requirements for Qwen/Qwen2.5-Math-72B-Instruct running on AMD Instinct MI250X.

Speed Snapshot

Topline estimate from compatibility data

ModelQwen/Qwen2.5-Math-72B-Instruct

GPUAMD Instinct MI250X

Q4 speed105 tok/s

Q4 VRAM required36GB

Data Source

Calculation and benchmark status

Speed values come from the compatibility dataset (`estimatedTokensPerSec`) and are sorted by quantization.

For full verdict logic and alternate GPUs, see the canonical compatibility page.

Quantization Speed Table

Quantization	VRAM needed	VRAM available	Speed	Verdict
Q4	36GB	128GB	105 tok/s	✅ Fits
Q8	71GB	128GB	66 tok/s	✅ Fits
FP16	142GB	128GB	35 tok/s	❌ Not recommended