// RUN โ€” FITS

Can the Intel Arc B580 12GB run Codestral Mamba?

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

Codestral Mamba on the Intel Arc B580 12GB

QuantFits?Max contextSpeedBeyond context
FP16 no โ€” โ€” โ€”
Q8 yes 17k 36.3-45.1 t/s ~3.2-4 (slow)
Q4 yes 42k 64.8-80.5 t/s ~5.7-7.1 (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 Arc B580 12GB run Codestral Mamba?โ–ถ

Yes. At Q4 it generates ~64.8-80.5 tok/s and fits up to 42k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~5.7-7.1 tok/s.

How much context fits?โ–ถ

At Q4, up to about 42k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.