$CRDO just got a lower price target but the underlying AI connectivity thesis remains intact.

Mizuho’s Vijay Rakesh maintained an Outperform rating on Credo but cut his price target from $290 to $245, citing lower peer multiples rather than a deterioration in Credo’s fundamentals.

The growth pipeline is still compelling. Amazon’s Trainium 3 ramp could drive roughly 50% higher cable volumes generation-over-generation, while SpaceX is targeting a major capacity expansion from roughly 2GW in 2026 to 7–10GW by the end of 2027.

Microsoft, Meta and Oracle add further potential demand, while Nvidia’s Blackwell platform could keep AEC volumes elevated into 2027. Rakesh also sees upside from 1.6T connectivity, optical DSPs, silicon photonics, Groq and Google opportunities.

The interesting part is that $CRDO sold off roughly 20% after earnings despite a substantial beat, as investors focused on margin pressure and the lack of an increase to the optical revenue target.

How much of that demand converts into revenue and margins as connectivity speeds move toward 200G and 1.6T.

$245 isn't $290, but Mizuho's thesis says the pullback could create another opportunity if Credo executes on the next wave of AI connectivity demand.