Direct:
Japan's diamond semiconductor sector is graduating from lab demos to actual facilities, and the timing finally looks credible. DigiTimes reports that university-backed startups are advancing factory builds, sample production, and device demonstrations for high-frequency, high-power, and harsh-environment applications. The parallel reporting backs this up across three distinct programs which is what makes the story worth exploring.
Ookuma Diamond Device is the lead actor. The Hokkaido University and AIST spinout broke ground in Okuma Town in March 2025, with completion and operations targeted for fiscal 2026 (April 2026 through March 2027). The initial application is diamond semiconductor amplifier and detector modules for neutron monitoring during fuel-debris removal at Fukushima Daiichi. That's one of the few real-world settings where high temperature, high radiation, and in-situ neutron sensing are not theoretical requirements. Ookuma has raised roughly ¥7.4 billion to date and claims over 90% yield at lab scale on a vertically integrated process from substrate to packaging. They're also in discussions with more than ten downstream customers in aerospace and telecom.
Power Diamond Systems, the Waseda University spinout, is on a different track. PDS demonstrated vertical diamond power MOSFETs handling hundreds of volts and signed a joint research agreement with JAXA in July 2025 to qualify the devices for space environments. Ground-based verification began in FY2025, with March 2026 device milestones reported at 550V/0.8A and 200V/1A switching. Commercial targets are 2030s, not now.
Orbray sits upstream as the substrate play. In February 2026 they reported producing a 50mm single-crystal CVD diamond wafer with Element Six at thermal conductivity above 2200 W/m·K, suitable for direct bonding to GaN. That addresses the wafer-scale bottleneck the industry has been stuck on. Mirise Technologies (Toyota/Denso-backed) is the automotive design partner; Orbray is targeting a 2029 IPO.
What this is not: a consumer storage story. Theoretical figures of merit put diamond well ahead of SiC and GaN, though the exact multiple depends on which metric and which assumptions, and theoretical is doing a lot of work in that sentence. Near-term applications are radar, satellite power, EV inverters, and nuclear instrumentation. The supply chain angle is that synthetic diamond is grown from methane and hydrogen rather than gallium-bearing inputs, which sidesteps the gallium export-control exposure that currently routes through China. That's why US defense primes are paying attention to the Element Six/Orbray work.
Worth watching, not worth hyping. Ookuma is nearest-term, PDS is 2030s, Orbray is the enabling infrastructure underneath both. The factory turning on by end-2026 is the next real checkpoint.
Sources
- DigiTimes, “Japan's diamond chip startups move toward production…” (May 12, 2026)
- DigiTimes, “Japanese startup pushes diamond semiconductors toward commercialization in EVs and satellites” (Dec 2025)
- PR Times / PDS device milestone announcement (Mar 2026)
- EE Times Japan, PDS coverage (Dec 2025)
- TechCrunch (sponsored content, Element Six), “Breaking the Thermal Barrier…” (Feb 2026)
- TechCrunch editorial, “ODD taps $27M for diamond chips…” (Oct 2024)
- JapanGov Kizuna, “Diamond Semiconductors: Turning Crisis into Innovation” (Feb 2025)
- Coral Capital, “Japan's Apollo Moment: Diamond Semiconductors” (Apr 2026)
- Ookuma Diamond Device corporate site (accessed 2025)
- Orbray press, 50mm wafer milestone with Element Six (Feb 2026); Orbray-MIRISE collaboration; Orbray 2029 IPO plan
- Nikkei Asia via kr-asia, diamond semiconductor R&D coverage (Sep 2024)
Reddit: https://ift.tt/x5CYdzM