// RUN โ€” FITS

Can the NVIDIA B200 run HiDream-I1?

Yes โ€” here's how fast and up to what context, with real measured numbers.

HiDream-I1 on the NVIDIA B200

QuantFits?Max contextSpeedBeyond context
FP16 yes 531k 162.2-193.3 t/s ~0.8-1 (slow)
Q8 yes 599k 324.4-386.7 t/s ~1.6-1.9 (slow)
Q4 yes 630k 579.4-690.5 t/s ~2.9-3.5 (slow)

Estimate (memory-bound). Beyond "max context" the KV cache spills to system RAM offload (depends on your system RAM) and speed drops to the "beyond" figure.

FAQ

Can the NVIDIA B200 run HiDream-I1?โ–ถ

Yes. At Q4 it generates ~579.4-690.5 tok/s and fits up to 630k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~2.9-3.5 tok/s.

How much context fits?โ–ถ

At Q4, up to about 630k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.