AI cluster scaling pulls optical interconnect demand forward, but supply, yields, and qualification govern who captures value. The 2031 upside case hinges on converting vertical integration into reliable volume and higher-margin mix, while reducing cyclicality via contracted capacity and trust features.

AI clusters make bandwidth-per-watt and qualified photonics supply the binding constraint; Coherent’s vertical InP stack + assembly scale + “trust/SLA” monetization can turn a cyclical components narrative into a durable interconnect infrastructure compounder by 2031.

Picks-and-shovels to AI compute scaling: optical I/O is a hard physical bottleneck, and manufacturing learning curves + supply assurance become the moat as cognition commoditizes.

Coherent’s credible non-linear upside comes from being on the constraint line of AI networking: (1) more optics content per AI cluster (800G→1.6T, higher lane counts, tighter power budgets), (2) persistent scarcity of qualified InP devices/modules that rewards vertically integrated supply, and (3) mix shift + learning-curve yields that expand gross margin while revenue scales. Additive opportunities (take-or-pay capacity + SLAs; authenticated/attested optics; OCS ecosystem) aim to convert “box margin” into stickier, higher-trust revenue streams, reducing cyclicality and protecting pricing when supply loosens.

The core risk is not “does AI need optics” but whether Coherent converts demand into qualified output at attractive yields while absorbing capex, working-capital swings, and customer pricing leverage. A second-order risk is policy/execution discontinuities (export controls, compliance outcomes, and ramp quality excursions) that can force abrupt resets in revenue/margin expectations.

Slightly higher multiple vs prior run reflects improved near-term demand visibility and clearer constraint-driven upside, partially offset by valuation already being elevated.

$203.47