Fabrinet will try to prove the 800G→1.6T ramp is shipping, not just selling design wins

Investors looking only at Lumentum or Coherent can miss the “manufacturing truth” of the 800G-to-1.6T transition—because those names can have lots of design activity without a clean read on how quickly components translate into end-customer shipments.

Fabrinet sits above the core photonics building blocks and below the hyperscaler branding: it packages and manufactures optical modules and datacom systems as a contract manufacturer/ODM. That structure makes its quarterly shipment volumes and margin profile a closer proxy for whether 1.6T capacity is moving from qualification to production.

Fabrinet already provides a useful filter for optical demand by splitting “Optical communications” into major end-markets (Telecom, Datacom, and Datacenter interconnect/DCI)—even though the company does not break those buckets down publicly into an explicit 800G vs. 1.6T unit count.

In its Form 10-Q for the quarter ended Dec 26, 2025, Fabrinet reported that “Optical communications” was $832.6M over the three months ended Dec 26, 2025, representing 73.5% of total revenues. It also listed Datacom as $278.1M and DCI as $142.2M for that same three-month period. This mix framing matters because 1.6T deployment is most likely to show up first in data center interconnect and datacom deployments rather than telecom.

Even without a public 800G/1.6T unit breakdown, you can still reason causally: contract manufacturing margin pressure tends to show up when ramps are happening but yields are immature; later, once volume stabilizes, gross margin typically normalizes. Working-capital movement can confirm whether revenue is backed by shipments.

A ramp can be messy, but Fabrinet has shown the ability to generate profitability at scale and maintain a liquid balance sheet. That means the market question is less about “can it fund production?” and more about “will production ramp cleanly enough to protect margin and cash conversion.”

If Fabrinet’s Q4 FY2026 shows an optical communications mix improvement and margin resilience, it supports a downstream interpretation: hyperscale datacom/DCI equipment demand is purchasing manufactured optics in volume.

Upstream, that kind of outcome typically increases confidence for photonics and module component supply (laser/photodiode/silicon-photonics packaging interfaces and the specialized electromechanical/optical assembly inputs that contract manufacturers rely on). Even though those upstream providers may not disclose 800G vs. 1.6T unit counts, the ODM layer’s throughput signal can still shift the market’s expectation for next-quarter build rates.