What is a Diamond Semiconductor — An Easy Guide to the "Ultimate Semiconductor"
When you hear the term "diamond semiconductor," you might first think of gemstones. But the diamond referred to here is not a ring stone — it is a "synthetic diamond" produced inside factory equipment by stacking carbon atoms layer by layer, using methane gas (the main component of natural gas) as the raw material. Using a technique called chemical vapor deposition (CVD), high-purity single-crystal diamond is created, then a small amount of impurities is introduced to allow electrical conduction, turning it into semiconductor devices such as transistors and diodes. In other words, a diamond semiconductor refers to diamond used as a "semiconductor material" in place of silicon or silicon carbide (SiC).
Why diamond? The history of semiconductors is also a history of seeking materials that can withstand ever harsher conditions. Silicon, which powers the brains of PCs and smartphones, is versatile — but in the world of power semiconductors, where "large amounts of power must be handled at high voltages," such as inverters that drive electric vehicle (EV) motors, power grids, and data center power supplies, silicon hits its limits in terms of heat and voltage. This is what drove the practical adoption of SiC and GaN (gallium nitride) from the 2010s onward — materials with a wider "band gap (the energy barrier that switches between conducting and non-conducting electricity)" than silicon. SiC is already used in EVs and railways, while GaN is used in fast chargers and telecommunications base stations. Diamond sits at the final destination in this lineage of "wide bandgap semiconductors," and because it theoretically surpasses SiC and GaN in performance, it is called the "ultimate semiconductor" and the "next-next-generation power semiconductor."
What specifically will change? Three scenarios are easy to visualize. The first is EVs. For the same level of performance, inverters can be made dramatically smaller and lighter, and cooling systems can be simplified, creating more range and interior space. The second is radar and satellite communications for defense and space. High-frequency radio waves can be amplified at high power, and the heat generated can be dissipated quickly, bringing closer to reality high-performance radars capable of detecting small threats like drones more rapidly, as well as communication devices that continue to operate under the intense radiation of outer space. The third is the very origin of this technology: nuclear reactor decommissioning sites. Diamond may make it possible to build sensors and circuits that operate in high-radiation environments — like those surrounding the debris (melted nuclear fuel) at the Fukushima Daiichi Nuclear Power Plant — where neither humans nor conventional electronics can approach.
Overwhelming Advantage from a Materials Perspective, and a Challenge That Has Persisted for 40 Years
The reason diamond is called the "ultimate" material becomes immediately clear when you line up the numbers for its physical properties. Its bandgap reaches approximately 5.5 eV—about five times that of silicon's roughly 1.1 electron volts (eV)—far surpassing even SiC (approximately 3.3 eV) and GaN (approximately 3.4 eV). Its dielectric breakdown field strength, which indicates how well a material resists electrical breakdown, is approximately 30 times that of silicon, allowing it to withstand higher voltages at the same thickness. Its thermal conductivity of approximately 2,000–2,200 W/m·K is the highest of any semiconductor material, equivalent to several times that of copper—meaning the semiconductor itself acts as an excellent heat sink. Carrier mobility is also high. Taking all of these factors together and comparing them using the "Baliga Figure of Merit (FOM)," which indicates the relative performance of power semiconductor materials, diamond is estimated to be more than 80 times that of SiC and more than 10 times that of GaN (according to technical reports from Asahi Kasei and others).
In plain terms, what these numbers mean is: "higher voltages, in smaller devices, even when running hotter, with lower losses." In theory, diamond holds the potential to make power converters built with silicon or SiC an order of magnitude smaller and lighter, while also dramatically reducing energy losses. This is precisely why the semiconductor industry has had its eye on diamond for more than 40 years.
Yet for a long time, practical application remained the quintessential example of a "seemingly achievable but never achieved" challenge. There are two major obstacles. The first is that doping—adding impurities to allow current to flow—is extremely difficult, particularly for stable "n-type" doping that carries electrons. The second is that while silicon can be produced cheaply in large discs (wafers) 300 mm in diameter, single-crystal diamond could only be grown in small pieces of at most a few millimeters to a dozen or so millimeters across, making it impossible to meet the size, uniformity, and low-cost requirements necessary for mass production. "A material that humanity spent forty years trying and failing to commercialize" (as Nippon.com put it)—this long-standing barrier has begun to fall, one breakthrough at a time, in the latter half of the 2020s, at the hands of Japanese research institutions and startups. That is the central subject of this article.
Why Japan Leads the World — 25 Years of Accumulated Expertise and the Culture of "すり合わせ" (Collaborative Adjustment)
Japan's prominence in diamond semiconductors is starkly evident in the sheer number of researchers in the field. Excluding China, there are only around 100 diamond semiconductor researchers in the entire world. Approximately a quarter of them are concentrated at a single Fukushima-based startup — Okuwa Diamond Device, which will be discussed later — and two of the most globally cited scientists in this field are also employed there. This is not an overnight achievement, but a reflection of decades of foundational research steadily built up with national funding by institutions such as the National Institute of Advanced Industrial Science and Technology (AIST), the National Institute for Materials Science (NIMS), Saga University, and Waseda University over the course of a quarter century.
James Riney, founding partner and CEO of Tokyo-based venture capital firm Coral Capital — an investor in the company — described this as "an advantage that could only have emerged from Japan" in his essay "Japan's Apollo Moment," published in April 2026. In his view, manufacturing diamond semiconductors is more of an "artisanal process" than a standardized mass-production workflow, requiring meticulous quality control at every stage from crystal growth to substrate processing to device fabrication. Japan's *monozukuri* (manufacturing) culture and skilled workforce are well-suited to this. In addition, the country is home to world-class power and analog semiconductor manufacturers — including Mitsubishi Electric, Fuji Electric, Toshiba, and ROHM — providing an industrial ecosystem capable of absorbing new materials. Riney writes that "25 years of accumulated knowledge is not something that well-funded competitors can replicate in a short period of time."
In practice, Japan's players can be broadly organized into four camps: AIST, serving as the command center that tackles core challenges in foundational technology; Okuwa Diamond Device, building the world's first mass-production facility; Orbray, leading in wafer (substrate) quality and large-diameter advancement; and Diamond Semiconductor (DSC), a Saga University spinout working on high-frequency devices. Adding Power Diamond Systems (PDS), a power device company spun out of Waseda University, gives a near-complete picture of Japan's lineup. Each will be examined in detail below.
From Fukushima: The World's First Mass Production Factory — Okuma Diamond Devices
A symbolic presence in this space is Okuma Diamond Device (headquartered in Sapporo), a company born from the technologies of Hokkaido University and AIST. Founder and CEO Naohisa Hoshikawa first visited Professor Junichi Kaneko's laboratory at Hokkaido University in 2016. Though he was already running another company, he sought out "technology with the scale to transform an entire industry" and made a career pivot. Over roughly six years, he taught himself physics, built trust with researchers, and in March 2022 co-founded the company alongside Kaneko (Hokkaido University) and Hitoshi Umezawa (AIST). The starting point was not a business opportunity but a social challenge — the 2011 Fukushima Daiichi nuclear accident. A national research project to build neutron detectors that could withstand proximity to nuclear debris became the prototype for this company. In September 2024, Okuma Diamond Device was also selected for Forbes Asia's "Forbes Asia 100 to Watch 2024."
What Okuma Diamond Device constructed in an industrial park in Okuma Town, Fukushima Prefecture, is the world's first mass-production factory for diamond semiconductors. The site covers approximately 5,800 square meters. On March 27, 2025, a groundbreaking ceremony was held with more than 70 attendees, including the town mayor, a Vice Minister of Economy, Trade and Industry, and representatives from TEPCO and Tohoku Electric Power. The factory building was completed and its inauguration ceremony held on May 29, 2026. However, completing the building and achieving full-scale mass production are two separate milestones; according to reports, following equipment installation and commissioning, the target for full operation (full-scale mass production) is set for fiscal year 2028. Production capacity will reach a maximum of several hundred thousand units annually. The initial target market is high-radiation environments such as decommissioning robots, with plans to expand applications into space, defense, telecommunications, and electric vehicles.
From a VC perspective, this company's strategy has a "quality and soundness unlike anything else." First, diamond is synthesized from methane gas and does not depend on rare earths or supply chains controlled by geopolitically adversarial nations. Coral Capital's Rainey describes this as "fully domestic production within a U.S. treaty-allied nation." Second, Rainey draws an analogy between this technology and the U.S. Apollo program. His argument runs: "The Apollo program was never really about the moon itself — breakthroughs in materials science, computing, communications, and manufacturing transformed civilian life. Diamond semiconductors, too, were born from the necessity of decommissioning Fukushima, and will find their way to markets the founders never imagined — defense radar, space, EVs." He also cites CEO Hoshikawa's engineer-minded caution — deliberately waiting six years "until commercialization could be done with integrity" — as a quality he can genuinely appreciate amid the hype-heavy world of deep-tech investment.
The Race Toward Larger Wafer Diameters — Orbray, EDP, and AIST
Whether diamond semiconductors can enter mass production ultimately hinges on "how large, high-quality wafers can be produced at low cost." Leading this charge is Orbray (formerly known as Adamant Namiki Precision Jewel), which has established a path to mass-producing 2-inch (approximately 50 mm diameter) diamond wafers and is advancing R&D on 4-inch substrates. The company has established production technology for the world's largest twin-free (111) single-crystal diamond freestanding substrates at 20 mm square with the crystal orientation suited for semiconductor devices, and is also working on the development of n-type diamond substrates — a key to practical application. In June 2024, Orbray partnered with Element Six, a subsidiary of De Beers, the world's largest diamond company, in the large-diameter, high-quality single-crystal synthetic diamond business. Orbray has also joined forces with MIRISE Technologies, a joint venture of Toyota Motor Corporation and DENSO, for research and development of diamond power devices, laying the groundwork for EV applications.
EDP Co., Ltd. (EDP), a manufacturer of synthetic diamonds, is also rapidly raising its profile. On May 27, 2026, the company announced that it had produced an approximately 53 mm square crystal by bonding four 25 mm-class single crystals using a "mosaic crystal" structure — a technique for laterally joining multiple single crystals — and achieved surface smoothness of approximately 5 nm (nanometers) across nearly the entire surface. By using this as a seed crystal and cutting it into circular shapes with a laser, mass production of 2-inch wafers is expected to become feasible, with mass production planned for the second half of fiscal year 2026. The company also intends to pursue 4-inch wafers in parallel, exploring single crystals of 50 mm square or larger, mosaics exceeding 100 mm square, and alternative bonding wafer technologies.
Leading the entire effort on the technical front is AIST (National Institute of Advanced Industrial Science and Technology). On February 2, 2026, AIST and EDP jointly announced a new method for bonding small diamond wafer chips (12 mm square) to silicon substrates (2-inch) at high temperatures of 1,200°C to suppress warping caused by differences in thermal expansion (published in the academic journal *ACS Applied Engineering Materials*). The height difference across the substrate, which was 27 μm at 1,000°C bonding, improved by approximately 60% to 9 μm at 1,200°C, and the interface reportedly withstood chemical treatment and heat treatment at 1,000°C. AIST has set a goal of realizing 6-inch wafers by 2030 and plans to accelerate the transfer of technology to domestic companies. The current landscape of wafer development is one of two parallel approaches: "growing larger crystals" (Orbray and EDP) and "connecting small pieces to create large areas" (AIST's bonding technology).
The Two-Way Player of the Device Faction——The Power-Oriented PDS, and Saga University's High-Frequency DSC
Device makers that fabricate actual components on wafers are broadly divided into two camps: "power" and "high-frequency."
The leading power player is Power Diamond Systems (PDS, established August 2022, headquartered in Shinjuku, Tokyo), founded on the technology of Professor Hiroshi Kawarada of Waseda University. Professor Kawarada is a world-renowned authority in the field, having developed a hydrogen-terminated channel in 1994, a silicon-oxide-terminated channel in 2020, and the world's first vertical diamond transistor. CEO Tatsuya Fujishima, who came from Rohm and went on to research GaN (gallium nitride) devices at the Massachusetts Institute of Technology (MIT), leads the company, while Professor Kawarada serves as co-founder and CSO (Chief Science Officer). Waseda University Ventures (WUV) invested 100 million yen at the time of founding. The company unveiled a diamond MOSFET (field-effect transistor) at SEMICON Japan 2025, confirmed continuous switching operation of a step-down DC-DC converter using a diamond MOSFET in March 2026, and announced a monolithic bidirectional switch in April that suppresses on-resistance to less than one-tenth of conventional structures — successively demonstrating 200V/1A-class switching operation. The target market is power conversion for applications such as EVs and base stations, where high voltage and high temperature performance matters.
The high-frequency segment is addressed by Diamond Semiconductor Co., Ltd. (DSC), a company grounded in the research of Professor Makoto Kasu and colleagues at Saga University. DSC was established on February 10, 2025 (with Kazuko Kasu as Representative Director) and was granted the designation of a Saga University-launched venture in June of the same year. Saga University partnered with JAXA (Japan Aerospace Exploration Agency) to fabricate T-shaped fine gate structures using electron beam lithography, achieving world-class radio wave amplification in the microwave/millimeter-wave band at 120 GHz. Manufacturing equipment is supplied by JEOL (Japan Electron Optics Laboratory), and collaborative research is also underway with JVC Kenwood. From January 2026, DSC began sample manufacturing and sales of what are claimed to be the world's first diamond semiconductor devices. The targets are high-frequency, high-output applications such as Beyond 5G/6G base stations and satellite communications, with a business model oriented toward high profitability through a fabless approach (no in-house fabrication) rather than chasing volume. With PDS handling "power for controlling electricity" and Saga University/DSC handling "high frequencies for amplifying radio waves," Japan is leading the world at the device layer as well.
Funding Sources and Money Flow — VC, State, and Defense Money
Following not just the technology but also the money reveals that Japan's diamond semiconductors are supported by three layers: VC, the state, and defense.
The funding trajectory of the core player, Ōkuma Diamond Device, has been staged. In its May 2023 seed round, Coral Capital led with ¥140 million, with Globis Capital Partners also participating. The subsequent pre-Series A round in October 2024 raised approximately ¥4 billion, including debt financing. Globis Capital Partners led the round, with Coral Capital, Green Coin Invest, Astart, Yucho Spiral Regional Innovation, Mitsui Sumitomo Insurance Capital, SMBC Venture Capital, FFG Venture Business Partners, Hokuho Capital, and Shinsei Corporate Investment joining as subscribers, while Mizuho Bank served as the lead debt financier. Including grants, total cumulative funding has reached approximately ¥6.7 billion. On the device side, WUV has invested ¥100 million in PDS, as noted above.
The weight of state and defense funding is also significant. According to Coral Capital, Ōkuma Diamond Device has received multi-year commissioned research contracts from Japan's Ministry of Defense — a backdrop shaped by the global surge in defense spending, with Japan's own defense budget roughly doubling over the past three years. Diamond semiconductors can amplify high-frequency signals in radar at high output with low heat generation, and as drop-in replacements for GaN modules they can boost performance on existing platforms without redesigning them, making them highly valuable for military applications. The simultaneous convergence of three funding streams — VC money (private growth capital), national research projects (long-term funding from basic research through to mass production), and defense procurement (high-certainty early demand) — is what underpins Japan's strength in this field from a financial perspective. It should also be noted that the uninterrupted, long-term flow of this public-private funding is itself the source of the "25 years of accumulated expertise" and depth of researcher talent described earlier.
How Silicon Valley and the World Are Reporting It — A VC Perspective
So how are Silicon Valley VCs and overseas media viewing this movement? To cut to the conclusion, the sharpest VC analysis has ironically come from Tokyo-based Coral Capital. The firm's CEO, James Riney, is an American investor who came from 500 Startups and knows the Silicon Valley playbook. His "Japan's Apollo Program" thesis frames diamond semiconductors in the exact vocabulary that today's Silicon Valley finds most compelling: deep tech + geopolitics + defense. His argument boils down to three points: (1) deep accumulated expertise in materials science combined with a mass-production culture; (2) a clean supply chain that requires no rare earths and has no dependence on adversarial nations; and (3) a technology born out of necessity (Fukushima) that is now reaching unexpectedly vast markets (defense, space, EVs).
Coverage by overseas media also grew more substantive entering 2026. Nikkei Asia reported that "Japan's R&D is bringing powerful diamond semiconductors closer to reality," while Nippon.com described the completed factory in Ōkuma Town as the world's first facility capable of producing "a material humanity has spent 40 years trying and failing to commercialize." Taiwan's industry publication DigiTimes ran a feature in May 2026 headlined "Japan's diamond chip startup moves toward mass production with factory and samples." The Japanese government's own communications site, JapanGov, has framed the story as "turning crisis into innovation with the ultimate deep tech," and international messaging around this theme has clearly been stepped up.
At the same time, it is worth soberly noting that pure Silicon Valley VC money has not been flooding into diamond semiconductors en masse. U.S.-side funding is moving primarily through defense frameworks. DARPA launched a "Ultra-Wide Bandgap Semiconductors (UWBGS)" program led by Element Six (a De Beers subsidiary), with an international team that includes Japan's Orbray (specialists in large-area diamond), radar giant Raytheon/RTX, France's Hiqute Diamond, and Stanford and Princeton universities. America's serious investment in diamond is defense-driven — and the fact that Japan's Orbray sits at its core is symbolic. On the private startup side, Advent Diamond, a spin-out from Arizona State University, has received grants including $750,000 from the National Science Foundation to develop diamond diodes and GaN-on-diamond. Adam Khan's AKHAN Semiconductor raised a cumulative $30.04 million across six rounds, but had its assets acquired by Diamond Technologies (DTI) in June 2025. San Francisco's Diamond Foundry has raised a cumulative $315 million and was valued at $1.8 billion in 2021, making it a notable player — but its focus is on lab-grown diamonds for jewelry and wafers for solar cells, and it is hard to call it a leading actor in power semiconductors. On the whole, "smart money" in Silicon Valley is converging on a division-of-labor thesis: Japan holds the lead in materials and talent, while the most reliable demand lies in U.S. defense.
Geopolitics and Supply Chains — The US-Japan $550 Billion Deal and the "Second Rare Earths"
What is propelling diamond semiconductors in a broader geopolitical context is the Japan-U.S. trade and investment agreement concluded in July 2025. Under this framework, the U.S. would apply a reciprocal tariff of 15% on Japanese goods in exchange for Japan committing to a total of $550 billion (approximately ¥85 trillion) in investment in the United States. On February 18, 2026 (Japan time), the two governments announced the first wave of three projects: approximately $33.3 billion (roughly ¥5.2 trillion) in gas-fired power generation for AI data centers (Ohio); approximately $2.1 billion (roughly ¥330 billion) in export infrastructure for U.S.-produced crude oil (Texas/Gulf Coast); and approximately $600 million (roughly ¥93 billion) in synthetic diamond manufacturing (Georgia).
It is worth noting that the synthetic diamond project in this first wave is strictly the production of industrial synthetic diamonds (grit — abrasive grains used in grinding and polishing), and is not a diamond semiconductor fabrication facility per se. The operator here is again Element Six (a De Beers subsidiary), with the stated aim of supplying materials for ultra-precision polishing and processing of automotive, aerospace, and semiconductor components, as well as quantum devices and military radar components, to meet domestic U.S. demand. That said, the materials are on a continuum with semiconductor-grade diamond, and above all, the logic of "breaking dependence on China" is identical to that for semiconductor applications. China holds a large share of global production of industrial synthetic diamonds — estimates vary across reports from over 60% to over 90% — with *Nikkei Business* citing "over 90%" and presenting concerns about this as "the second rare earth."
That said, there are caveats regarding how much this investment actually serves Japan's economic security. As *Nikkei Business* and the Nomura Research Institute have pointed out, there is no explicit provision guaranteeing priority supply of U.S.-manufactured diamonds to Japan; the U.S. Department of Commerce's stated objective is limited to "meeting domestic U.S. demand." Japanese companies such as Asahi Diamond Industrial and Noritake have been mentioned as potential buyers, and the president of the Japan Bank for International Cooperation (JBIC) described the project as "mutually beneficial and sufficiently bankable," yet criticism also persists that "the security dividend for Japan is hard to see." This is a point that should be understood with an awareness of the reporting inconsistencies and underlying uncertainty. In any case, the fact that diamond — not as a gem or an abrasive, but as a strategic material — has made it onto the table in Japan-U.S. trade negotiations speaks to the growing weight of this field.
Future Roadmap and Market Size——When and What Will Happen
Market size forecasts vary widely depending on how the market is defined. For "diamond semiconductor substrates" alone, some estimates point to growth from approximately $420 million (about ¥65 billion) in 2024 to roughly $790 million (about ¥120 billion, ~11% CAGR) by 2030. When broadened to "diamond materials for semiconductors" — including heat dissipation materials — other projections see expansion from around $1.5 billion (about ¥230 billion) in 2023 to approximately $3.7 billion (about ¥580 billion, ~12% CAGR) by 2030. More bullish research posits growth from roughly $2.1 billion (about ¥330 billion) in 2025 to approximately $17.8 billion (about ¥2.8 trillion, ~26% CAGR) by 2034. Since the magnitude of these figures depends entirely on "what is included in the market," it is inadvisable to take any single firm's forecast at face value; the appropriate framing is that this is "an early-stage market poised for double-digit annual growth."
Looking at the timeline, key milestones are coming into sharper focus. The year 2026 marks an inflection point from proof-of-concept to the start of supply: the completion of the Okuma Town factory (May), the world's first device sample shipments by Saga University and DSC (January), and EDP's establishment of 2-inch wafer mass production capacity (second half of the fiscal year). In 2027–2028, full-scale mass production at Okuma Diamond Device (target: FY2028) and the practical realization of 4-inch wafers come into view. Around 2030, AIST's target of achieving 6-inch wafers is expected to coincide with the full-scale adoption of diamond semiconductors in high-value-added sectors such as defense, space, and 6G. By 2030–2035, expansion into volume markets such as EV inverters and industrial power conversion is expected to begin — this is the roadmap each stakeholder is drawing.
From the perspective of investors and businesses asking "what should be monitored going forward," the key indicators are clear. First, wafer diameter (2-inch → 4-inch → 6-inch) along with yield and cost; second, breaking through the longstanding challenge of stable n-type diamond formation; third, initial orders from defense and space sectors and the first commercial design wins; and fourth, the extent to which national-level funding — such as the Japan-U.S. $550 billion investment framework — flows through to semiconductor-grade material supply. With each of these milestones achieved, the credibility of diamond semiconductors as a practical technology will rise sharply.
Summary — Can "Unbreakable" Materials Become Japan's Next Export Industry?
Diamond semiconductors are the "ultimate semiconductor" capable of directly breaking through silicon's physical limits — and Japan is unquestionably at the forefront of their commercialization. Ookuma Diamond Devices, which houses a quarter of the roughly 100 researchers worldwide in this field; Orbray and EDP, which continue to scale up wafer sizes; AIST, the strategic command center of the technology; and PDS for power applications and Saga University/DSC for high-frequency use — a Japanese contingent with clearly divided roles is transforming "laboratory dreams" into "factory reality," built on a quarter-century of accumulated expertise.
Viewed through the lens of a Silicon Valley VC, the essence of this story comes down to three things: a moat in materials and talent that only Japan can replicate; a clean supply chain requiring no rare earths and independent of adversarial nations; and an "Apollo Program"-scale ambition born out of Fukushima's necessity, now aimed at the massive markets of defense, space, and EVs. Funding flows from three layers — private VC, the state, and defense — and geopolitics, via the U.S.-Japan investment agreement, has elevated this material to the status of strategic asset. The remaining hurdles are larger-diameter wafers, n-type diamond, and mass-production costs, but between 2026 and 2030, answers will emerge for each of them, one by one. Japan's worst disaster, which struck in Fukushima, may prove to be the starting point for the country's next great export industry — and the years that will determine whether that hypothesis becomes reality have now begun.