The upside case depends on turning early AI connectivity orders into predictable, high-yield volume shipments across multiple customers. If execution clears that bar, a durable revenue ramp and a more normal hardware multiple become plausible by 2031.

POET is a “proof-of-repeatable-volume” call option on AI interconnect scaling: if its Optical Interposer platform converts 2026 pilot orders into high-yield, high-cadence shipments across multiple customers, it can earn a durable component position in next-gen datacenter links and grow into a credible merchant photonics platform by 2031.

AI clusters are bandwidth- and power-constrained; photonics that cuts cost/power/assembly friction is directly pulled forward by the compute buildout, even if POET isn’t a compute owner.

The non-linear upside is less about inventing new demand and more about crossing the manufacturability threshold that turns a promising photonics architecture into a repeatable “ship-it-every-week” supplier. If POET demonstrates stable yields, clear reliability data, and multi-customer pull in mainstream module form factors (e.g., OSFP), it can transition from bespoke programs to a platform play with better pricing power and lower per-customer engineering drag. In that state, the market typically awards a higher revenue multiple than an R&D-stage supplier, even if capex stays meaningful.

POET’s main risk is commercialization: qualification timing, field reliability, and sustaining high first-pass yield at volume. Second-order risks are customer/program concentration, pricing pressure as mainstream modules commoditize, and dilution/overhang from warrants if ramps require more capital than expected.

Updated dilution/overhang visibility and a more conservative risk-adjusted terminal multiple tempered the upside versus prior run.

$7.80