Today’s valuation is mostly option value on two gates: (1) customer acceptance of outsourced manufacturing routes and (2) conversion of initial orders into repeat POs with acceptable yields/field reliability. If those gates clear, optics demand in AI clusters can scale faster than typical hardware cycles, and POET can ride refreshes from
800G-class to higher-speed platforms with reuse of its integration approach. The 5-year case assumes POET becomes a niche-but-real supplier into AI
optical engines/light sources, not a dominant module vendor, and remains discounted vs larger peers due to scale and customer concentration.