Toshiba plans to double HDD production capacity for AI data centers within FY2027 from its 2025 level, targeting a 30% share by storage capacity, up from ~10%
Context & Ripple Effects
Toshiba’s storage expansion extends a broader infrastructure-capacity posture: after its 2023 plan to double power-management-chip output, the company is again tying manufacturing scale to data-center demand. The surrounding buildout includes Taiwanese AI-server shipments projected to double in 2025, making storage capacity a complementary constraint rather than a standalone market.
The move also lands amid a wider memory upswing, with Kioxia forecasting roughly 20% annual NAND-demand growth on AI demand. Toshiba is pursuing a distinct HDD-capacity position within that broader storage cycle, rather than relying only on semiconductor components.
First-order effects
- Toshiba plans to double HDD production capacity by fiscal 2027, directing more manufacturing capacity toward AI data-center storage and seeking to lift its share of installed storage capacity from about 10% to 30%.
- Data-center buyers will have a prospective larger Toshiba source for high-capacity HDD supply as they provision storage alongside AI-server deployments.
Second-order effects
- The plan raises the competitive importance of manufacturing scale in enterprise HDD supply, as Toshiba’s share target requires it to win capacity deployments rather than merely add output.
- The HDD investment broadens the hardware beneficiaries of AI infrastructure spending beyond flash suppliers such as Kioxia and server makers, transmitting demand into long-duration data storage.
Third-order effects
- If AI-server deployment continues at the pace indicated by prior coverage, storage capacity is likely to become a more durable leg of infrastructure spending, linking compute expansion to repeated investment in data retention hardware.
- Toshiba’s target points to an AI infrastructure supply chain in which component makers compete for share of total deployed capacity, not only unit shipments.
The trend: AI infrastructure demand is propagating from accelerators and servers into the storage systems needed to retain and serve expanding data workloads.