Ordered by predicted speed. Bandwidth sets decode rate; memory decides whether it runs at all.
Every figure here is computed, not measured. Fit is weights at each quantisation against device memory, with a 25% allowance for the KV cache, activations and the OS. Decode rate is the memory-bandwidth roofline at 70% efficiency. Nobody has run this model on these devices.
Device
Memory
Bandwidth
Best quality
Weights
tok/s
Fastest
NVIDIA GB200 Grace Blackwell Superchip
372.0 GB
16000.0 GB/s
q6
232.34 GB
~927.7
~1391.6 q4
AMD Instinct MI355X
288.0 GB
8000.0 GB/s
q5
193.62 GB
~556.6
~695.8 q4
AMD Instinct MI325X
256.0 GB
6000.0 GB/s
q4
154.89 GB
~521.8
~521.8 q4
Apple M3 Ultra
512.0 GB
819.0 GB/s
fp8
309.79 GB
~35.6
~71.2 q4
Where it runs
Taken from the card's availability section.
Platform
Id on that platform
Hugging Face
XiaomiMiMo/MiMo-V2-Flash
Verified benchmark evidence
Each row was checked against its source by a reviewer, and carries the model identifier as actually evaluated — which is not always the same as this card's.