AMD Instinct MI250X~104 tok/s (Q4)

nvidia/Llama-3.1-Nemotron-70B-Instruct-HF speed on AMD Instinct MI250X

Quantization-specific throughput and VRAM requirements for nvidia/Llama-3.1-Nemotron-70B-Instruct-HF running on AMD Instinct MI250X.

Speed Snapshot

Topline estimate from compatibility data

Modelnvidia/Llama-3.1-Nemotron-70B-Instruct-HF

GPUAMD Instinct MI250X

Q4 speed104 tok/s

Q4 VRAM required34GB

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	34GB	128GB	104 tok/s	✅ Fits
Q8	69GB	128GB	61 tok/s	✅ Fits
FP16	138GB	128GB	35 tok/s	❌ Not recommended