Toshiba Starts Shipping Test Samples of 1200V Trench-Gate SiC MOSFET that will Enhance Efficiency in Next-Generation AI Data Centers

· Source: The AI Journal · Field: Technology & Digital — Emerging Technologies & Innovation, Cloud Computing & IT Infrastructure · Depth: Advanced, short

Summary

Toshiba Electronic Devices & Storage Corporation has begun shipping test samples of "TW007D120E," a 1200V trench-gate SiC MOSFET. This component is primarily designed for power supply systems in next-generation AI data centers and is also suitable for renewable energy equipment. Addressing the increasing power consumption from generative AI and 800V high-voltage direct current (HVDC) architectures, the TW007D120E features Toshiba's proprietary trench-gate structure. This design achieves industry-leading low On-resistance per unit area (RDS(on)A), reducing conduction loss by approximately 58% and improving the figure of merit (On-resistance × gate-drain charge) by about 52% compared to Toshiba's 3rd-generation SiC MOSFET (TW015Z120C). Housed in a QDPAK package supporting top-side cooling, it enhances power density and thermal performance, contributing to lower power consumption, miniaturization, and higher efficiency. Mass production is planned for fiscal year 2026.

Key takeaway

For AI Hardware Engineers designing power supply systems for next-generation AI data centers or renewable energy, Toshiba's TW007D120E 1200V trench-gate SiC MOSFET offers substantial efficiency gains. Its reduced conduction and switching losses, coupled with a high thermal performance QDPAK package, directly address the demands of high-power AI servers and 800V HVDC architectures. You should evaluate test samples of this component to enhance power density, reduce heat generation, and improve overall system efficiency in your upcoming designs, especially as mass production approaches in fiscal year 2026.

Key insights

Toshiba's new SiC MOSFET significantly boosts power efficiency and density for AI data centers and renewable energy.

Principles

In practice

Topics

Best for: CTO, VP of Engineering/Data, AI Hardware Engineer, AI Architect, IT Professional

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The AI Journal.