// RUN โ€” FITS

Can the Intel Gaudi 3 run Zidong Taichu?

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

Zidong Taichu on the Intel Gaudi 3

QuantFits?Max contextSpeedBeyond context
FP16 yes 1676k 335.3-416.3 t/s ~3.6-4.5 (slow)
Q8 yes 1725k 670.6-832.5 t/s ~7.3-9 (slow)
Q4 yes 1747k 1197.5-1486.6 t/s ~12.9-16.1 (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 Intel Gaudi 3 run Zidong Taichu?โ–ถ

Yes. At Q4 it generates ~1197.5-1486.6 tok/s and fits up to 1747k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~12.9-16.1 tok/s.

How much context fits?โ–ถ

At Q4, up to about 1747k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.