AI supply-chain thesis — mapping bottlenecks, focus companies, and supply-chain exposure for investors.
**Bottleneck theme:** Storage **Focus:** $STX — Seagate Technology Holdings plc Mass-capacity HDDs are the AI cold-storage and training-data layer — the unsexy commodity tier holding the petabyte-scale data lakes that frontier models train on, plus model checkpoints, inference logs, and the cold-tier of vector databases. Seagate is the duopoly leader (with $WDC) in nearline HDD and the most aggressive HAMR ramp story in the industry: 30TB shipping today, 36-50TB on the technology roadmap, with hyperscaler qualifications well advanced. Lead times have stretched into 2027, prices are firming, and supply discipline is tighter than at any point since 2018. The thesis is dual-engine: cyclical recovery from the 2022-2023 inventory correction, plus a secular re-rating as the market finally values nearline HDD as core AI infrastructure rather than a melting ice cube. Higher-capacity HAMR drives also carry better margins than legacy CMR SKUs, so even flat unit shipments deliver meaningful EPS leverage. Pair with $WDC for duopoly exposure or own as the higher-conviction technology leader (HAMR is two years ahead of $WDC's roadmap).
Browse all AI supply-chain theses · Macro trends · Industries · Product categories