DB HiTek má kvalifikaci spolehlivosti pro osmipalcovou výrobu SiC MOSFETů s napětím 1 200 V

DB HiTek completed reliability qualification for its 1,200V SiC MOSFET process on 8-inch (200mm) wafers, positioning it to offer what it describes as the world's first complete 8-inch SiC foundry process. The company plans volume production in 2027, with broader customer access from Q2 2027; third-generation PDK availability is targeted for November 2026. Its second-generation process delivers on-resistance of 2.5 mΩ·cm² or below, while the third generation targets 2.3 mΩ·cm² or below and could reduce customers' product-development cycles by more than one year.

Analysis

The strategic value is not the process announcement itself but the potential shift from a captive-device SiC market toward a merchant-foundry model. If DB HiTek (000990 KS) can offer credible 200 mm capacity, fabless power-semiconductor designers gain an alternative to buying finished devices from vertically integrated suppliers such as onsemi (ON), STMicroelectronics (STM), Infineon (IFX GR), and Rohm (6963 JP). That could pressure device ASPs and reduce the value of integrated suppliers' capacity scarcity premium, but only after automotive-grade customer qualifications—typically a 12-24 month cycle—rather than in the next quarter.

The key economic uncertainty is yield, not electrical specifications. SiC wafer defect density, epitaxy consistency, and module-level reliability determine whether 200 mm lowers cost per die or merely increases scrap exposure; no yield, utilization, customer-design-win, or capacity data were disclosed. The announced development tools may accelerate prototype activity, but prototype demand is not evidence of recurring wafer volume, particularly while EV traction-inverter growth has softened and SiC inventory digestion remains a near-term industry risk.

Over the next 1-3 months, the relevant catalyst is third-generation process qualification and evidence of named customer tape-outs. Over 6-18 months, initial production ramps could be incrementally negative for pure-play merchant SiC pricing, especially Wolfspeed (WOLF), whose valuation relies heavily on 200 mm scale and a recovery in factory utilization. Contrarian view: investors may overread this as immediate competitive displacement; automotive customers prioritize proven field reliability and secured substrate supply, leaving incumbents' qualification moat intact unless DB HiTek demonstrates sustained yields and multi-year customer commitments.

Key Decisions for Investors

• No immediate directional position in 000990 KS: treat the release as a watch item until management discloses 200 mm SiC capacity, wafer yields, customer commitments, and expected 2027 revenue contribution. A re-rating without these metrics would be narrative-driven rather than earnings-supported.
• Maintain a 6-12 month relative-value watch: short WOLF / long ON or STM only if DB HiTek announces binding foundry volume with an automotive or industrial customer. The thesis is that incremental merchant supply undermines WOLF's scarcity premium, while ON and STM retain module, systems, and customer-qualification advantages. Falsifier: WOLF demonstrates materially improving utilization, positive gross-margin trajectory, or long-term take-or-pay wafer commitments.
• For IFX GR, STM, and ON, use any near-term SiC-supply-disruption narrative as an opportunity to avoid chasing upside: the longer-term risk is margin normalization in discrete SiC MOSFETs, while power-module integration and software/control content should remain more defensible than die sales.
• Set alerts for November process-qualification results and 2027 customer-access milestones. Upgrade the competitive threat only if disclosed reliability/yield data support cost-per-die parity and if customer tape-outs convert into production awards; absent both, the impact on listed SiC suppliers is likely immaterial through 2027.