Spoločnosť DB HiTek dokončila kvalifikáciu spoľahlivosti pre výrobný proces 8-palcových 1 200 V kremíkovo-karbidových tranzistorov typu MOSFET

DB HiTek completed reliability qualification for its 8-inch, 1,200V silicon-carbide MOSFET process and plans mass production in 2027, positioning it in the emerging 8-inch SiC foundry market. The company says it has established the world's first end-to-end foundry production flow for these devices and will release a third-generation PDK in November. Its second-generation process delivers on-resistance of 2.5 mΩ·cm² or less, while the third generation targets 2.3 mΩ·cm² or less and at least 2.5 μs short-circuit withstand time; DB HiTek expects the PDKs to reduce customer product-development cycles by more than one year.

Analysis

The strategic value is not the process milestone itself but the potential shift from vertically integrated SiC supply toward a merchant-foundry model. If DB HiTek can convert PDK access into meaningful customer tape-outs, it lowers entry barriers for power-semiconductor designers and could pressure the device-level pricing and captive-capacity advantage enjoyed by Wolfspeed (WOLF), STMicroelectronics (STM) and onsemi (ON). The effect is likely negligible for 2026 estimates, but becomes relevant to 2027-28 gross-margin assumptions if customers qualify a second source rather than merely use DB HiTek for prototyping.

The key bottleneck is yield, not stated electrical performance. SiC economics on 200 mm wafers only improve meaningfully when defect density, wafer availability, cycle time and module-level qualification all scale simultaneously; a reliability-qualified process does not establish those variables. Automotive customers also require extended PPAP and multi-quarter field-validation cycles, so an announced development-time reduction may accelerate design starts without generating material production revenue for 12-24 months.

Consensus may overstate the threat to established suppliers: large EV and industrial OEMs value bankable wafer supply, module integration and proven field reliability more than a modest die-cost advantage. Conversely, the development is incrementally negative for WOLF because its valuation and refinancing narrative depend on 200 mm manufacturing leadership translating into scarcity rents. Watch SiC device ASPs, 200 mm wafer yields, customer design-win disclosures and WOLF/STM/ON margin guidance; stable pricing and no external customer commitments by mid-2027 would falsify the competitive-disruption thesis.

Key Decisions for Investors

• No immediate directional trade on DB HiTek (KRX: 000990): treat the November PDK release and disclosed customer tape-outs as watch catalysts, not revenue evidence. Upgrade only if the company provides contracted wafer volumes, yield data or capex-backed 2027 utilization guidance.
• Maintain a 6-18 month relative-value watch: short WOLF versus long STM or ON if SiC pricing softens or new outsourced 200 mm capacity is contracted. The thesis targets multiple compression in the more execution- and balance-sheet-sensitive pure play; exit if WOLF demonstrates sustained yield/ramp progress and secures incremental long-term take-or-pay commitments.
• For existing STM and ON longs, do not reduce positions solely on this announcement; their near-term earnings are more sensitive to EV/inverter demand and inventory normalization than a prospective foundry entrant. Reassess on 2027 SiC gross-margin guide changes or evidence that customers are dual-sourcing externally.
• Set an alert around 2027 industry commentary on 200 mm SiC utilization and device ASPs. A simultaneous decline in utilization and ASPs would favor a broader underweight in SiC-exposed suppliers, while firm ASPs despite capacity additions would indicate qualification barriers remain the dominant constraint.