Instinct MI100 x4 vs RX 7900 XTX

GPU AGPU B

Best verified output tok/s per model (single-stream generation). Hover a value for the quant.

Verified data on this pair is thin — most of the board is still pending review. Submit your own run and help fill the matrix: one prompt, three steps, no downloads.

Instinct MI100 x41 verified run · best 132.4 tok/s

RX 7900 XTX6 verified runs · best 550.2 tok/s

Model	Instinct MI100 x4	RX 7900 XTX
Qwen3.5-0.8B	—	550.2 MQ4
LFM2.5-350M	—	398.4 MQ4
Qwen3.8-27B	132.4 INT8	154.2 MQ4
Qwen3.6-35B-A3B	—	150.0 MQ4R
Qwen3.6-35B-A3B-abliterated-v4	—	82.1 Q8_0

Head-to-head (models tested on both): Instinct MI100 x4 0 · RX 7900 XTX 1 · tied 0 · data as of Sep 23, 2026. Full rules: methodology.

Related reading: RTX 5090 vs 3090 · quantization · Value index · Full board