Analyst started position in CRDO at ~$192 (after 25% earnings pop) with high confidence in re-rating higher due to DustPhotonics vertical integration enabling CPO/NPO expansion; believes market hasn't priced in SiPho/CPO ramp yet.
DustPhotonics SiPho in-house gives full stack ownership (SerDes IP to system integration); insulated from EML laser shortages vs LITE; deployed in hyperscale AI + designed for NPO/CPO; FY2027 optical >$500M target; path to 3.2T where CPO mandatory.
$CRDO up 25% since earnings release: And high confidence in Credo re-rating higher alongside CPO TAM expansion. By bringing SiPho in-house via the DustPhotonics acquisition, resulting in vertical integration where they own: SerDes IP -> DSP -> SiPho PIC -> system integration -> telemetry. All resulting in TAM expansion for: transceivers + LPO + CPO. And since Credo's built on reliability (ZeroFlap, PILOT telemetry), that gives me more confidence that they'll do the same for CPO where the failure domain is much larger. Especially because they're insulated from EML laser shortages that $LITE have rn: Since DustPhotonics' L3C integrates optical functions onto a single PIC which uses CW lasers instead of scarce EMLs. In turn: Reduces laser count -> lowers thermal load -> improves yield -> cuts BOM cost at scale as port speeds advance beyond 800G. Plus, it's super positive that DustPhotonics' SiPho PICs are already deployed in transceivers at hyperscale AI clusters + already in design for NPO & CPO. - At OFC 2026 they demonstrated the full portfolio - 400G/800G ZeroFlap transceivers, 800G/1.6T optical DSPs. - With a path to 3.2T where CPO becomes mandatory. Meaning that Credo are expanding share of wallet w/ hyperscalers that already trust its reliability via AECs & ZeroFlap. For a rough catalyst path: -> FY2027 optical tracking to >$500M (H2 inflection) -> NPO/CPO design wins (SiPho earnings) -> 1.6T ramp and 3.2T roadmap progress -> If scale-up goes optical on $NVDA NVL576 -> NVL1152 cadence (2027-2028), Credo's optical-engine TAM inflects. Lots of moving parts with Credo + important to remember that CPO is for 2027 onwards. And that the FY2027 optical number (>$500M) is mainly transceivers + DSPs (800G-led), with SiPho contributing via in-house PIC integration, not large standalone PIC revenue. But to me, it seems like the market hasn't begun pricing in SiPho/CPO ramps yet.
| Horizon | Raw move | vs SPY |
|---|---|---|
| 1 day | +3.8% | +5.4% |
| 1 week | +6.5% | +4.7% |
| 1 month | -4.4% | -6.4% |
| 3 months | — | — |
| 6 months | — | — |