// RUN โ€” FITS

Can the NVIDIA B200 run BART-IT?

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

BART-IT on the NVIDIA B200

QuantFits?Max contextSpeedBeyond context
FP16 yes 2668k 20857.1-24857.1 t/s ~104.3-124.3 (slow)
Q8 yes 2670k 41714.3-49714.3 t/s ~208.6-248.6 (slow)
Q4 yes 2671k 74489.8-88775.5 t/s ~372.4-443.9 (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 BART-IT?โ–ถ

Yes. At Q4 it generates ~74489.8-88775.5 tok/s and fits up to 2671k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~372.4-443.9 tok/s.

How much context fits?โ–ถ

At Q4, up to about 2671k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.