// RUN โ€” FITS

Can the Intel Gaudi 3 run Cohere Command A Reasoning?

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

Cohere Command A Reasoning on the Intel Gaudi 3

QuantFits?Max contextSpeedBeyond context
FP16 no โ€” โ€” โ€”
Q8 yes 16k 19.3-24 t/s ~0.2-0.3 (slow)
Q4 yes 165k 34.5-42.9 t/s ~0.4-0.5 (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 Intel Gaudi 3 run Cohere Command A Reasoning?โ–ถ

Yes. At Q4 it generates ~34.5-42.9 tok/s and fits up to 165k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~0.4-0.5 tok/s.

How much context fits?โ–ถ

At Q4, up to about 165k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.