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. This is a mixture-of-experts model and no card carries active-parameter counts, so these predictions use total parameters and understate the real speed.
Device
Memory
Bandwidth
Best quality
Weights
tok/s
Fastest
NVIDIA GB200 Grace Blackwell Superchip
372.0 GB
16000.0 GB/s
q6
224.09 GB
~50.0
~75.0 q4
AMD Instinct MI355X
288.0 GB
8000.0 GB/s
q5
186.74 GB
~30.0
~37.5 q4
AMD Instinct MI325X
256.0 GB
6000.0 GB/s
q5
186.74 GB
~22.5
~28.1 q4
Apple M3 Ultra
512.0 GB
819.0 GB/s
fp8
298.79 GB
~1.9
~3.8 q4
Where it runs
Taken from the card's availability section.
Platform
Id on that platform
Hugging Face
tencent/Hy3
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.