The opportunity is to convert AI-driven bandwidth growth into durable content gains while using specialty manufacturing to win trusted, long-life sockets. The key debate is whether new adjacencies scale without triggering a cyclical derating.

MACOM can compound faster than “normal semis” by riding AI connectivity upgrades (800G→1.6T optics) while using its owned specialty manufacturing + defense-grade qualification to expand into higher-value sockets (semi-custom, assured supply) and selectively adjacent AI power (GaN), sustaining premium growth even as the cycle normalizes.

Hard-physics connectivity (optics/RF/power) becomes scarcer as AI scale compresses time-to-scale; MACOM’s moat is qualified sockets + process control, not attention/ads.

By 2030, the upside case is less about “more SKUs” and more about content-per-system rising in AI networks and power delivery, where qualification + reliability matter. If MACOM executes, it can keep taking share in high-speed optics building blocks and convert manufacturing control into assured-supply wins, while adding a credible GaN power line for AI racks. Multiple likely compresses versus today, but durable AI/defense mix can still support a premium to most analog peers.

The main risk stack is (1) AI/datacenter demand lumpiness plus customer concentration, (2) competitive/vertical-integration pressure in optics and emerging AI power, and (3) capital intensity/utilization risk from owned specialty manufacturing. Policy/export-control shocks can amplify both demand volatility and cost structure.

$189.13