The upside case depends on turning early production orders into repeatable shipments and expanding qualified manufacturing. The downside is that optics sockets commoditize before the company proves volume economics.

POET is a non-linear “qualification-to-volume” optics bet: if 2026 proves repeatable shipments from its Malaysia-centered manufacturing stack, it can scale from NRE-grade revenue to meaningful AI interconnect component volume by 2031, with upside amplified by platform-style packaging reuse across speed generations and customers.

AI drives a bandwidth arms race; POET’s core product sells “shovels” (optical connectivity) and could compound once qualified across multiple module makers, but it lacks hyperscaler-scale distribution moats.

The setup is convex: optics ramps are slow until a customer accepts qualification + supply chain, then orders can scale fast with each platform refresh. POET’s unusually strong cash position for its size reduces near-term financing fragility and lets it buy time to stabilize yields, qualify more lines, and pursue selective M&A. By 2031, a mid-single-digit share of the AI optics “engine” value pool is plausible if it wins multiple module-maker programs and keeps the interposer advantage in size/power/cost.

POET’s outcome is dominated by two binding gates: (1) qualification-to-repeat-order conversion, and (2) concentrated, qualification-locked manufacturing/supply dependencies. If either gate slips, the company can burn cash and dilute before reaching the scale where unit costs and margins become competitive versus entrenched optics supply chains.

$6.93