Kara

Diamond is the best thermal conductor known — 2,400 W/m·K, five times better than copper, and orders of magnitude beyond silicon carbide or gallium nitride. It also has a bandgap of 5.47 eV, a breakdown field of 10 MV/cm, and the highest electron and hole mobility of any wide-bandgap semiconductor. For decades, that combination has been a tease: a material that could solve thermal bottlenecks in power electronics, enable room-temperature quantum sensing, and shrink photonic devices — if only it could be grown reliably, in useful sizes, at a price engineers could justify.

Kara, a Y Combinator-backed startup out of Berkeley, is trying to close that gap. But its pitch is not another claim of bigger, purer, cheaper diamond. Instead, Kara is selling something more unusual: diamond grown to specification. The company's website leads with a table of material properties, a chart normalizing diamond against silicon, SiC, and GaN, and a promise that every wafer comes with a specified orientation, surface roughness, impurity profile, and isotope fraction. That is not the language of a jewelry maker or a lab-grown diamond novelty. It is the language of a semiconductor supplier.

The question is whether the market is ready to buy diamond the way it buys silicon — and whether Kara's spec-first approach is the right way to make that happen.

The thermal conductivity trap

Thermal management is the most immediate application for diamond, and the easiest to understand. As power electronics shrink and current densities rise, heat becomes the limiting factor. Silicon carbide and gallium nitride can switch faster and handle higher voltages than silicon, but they conduct heat poorly — SiC at 490 W/m·K, GaN at just 130 W/m·K. That means a GaN device can generate more heat in a smaller area than the substrate can safely dissipate, forcing engineers to add bulky cooling systems or derate the device.

Diamond's thermal conductivity is 2,400 W/m·K, roughly 18 times that of GaN and 5 times that of SiC. A diamond heat spreader or submount can pull heat away from a hot spot before it builds up, potentially enabling higher power density, longer lifetimes, or simpler cooling. That is the promise Kara is selling with its "thermal materials" category — heat spreaders and submounts for high-power devices.

The trap is that thermal management is a commodity market. Copper and aluminum are cheap, and even silicon carbide heat spreaders are well established. Diamond only wins when the heat flux is extreme enough that no other material can do the job — which is exactly the niche Kara is targeting, but it is a niche that requires proof, not just a datasheet.

Six properties, one material

Kara's website makes a deliberate argument: diamond is not just a thermal material, it is a multi-property platform. The company's chart normalizes six properties — bandgap, breakdown field, thermal conductivity, hole mobility, electron mobility, and saturation velocity — against the best-performing material for each. Diamond scores 100% on all six. Silicon, by contrast, scores 20% on bandgap, 3% on breakdown field, 6% on thermal conductivity, and 31% on electron mobility. GaN scores 62% on bandgap, 33% on breakdown, 5% on thermal conductivity, and 27% on electron mobility.

That is a striking visual: a material that wins on every axis. But it also raises the question of what to do with that breadth. Diamond is not going to replace silicon in logic chips — it is too expensive and too hard to dope. Instead, Kara is positioning diamond for applications where its combination of properties is uniquely valuable: power electronics, where high breakdown field and thermal conductivity matter; quantum sensing, where nitrogen-vacancy centers in diamond can operate at room temperature; photonics, where thin-film diamond can confine light in waveguides; and quantum networks, where spin-photon interfaces require precise isotopic control.

The breadth is both a strength and a risk. A startup that promises to serve six different markets may struggle to focus. But Kara's material platform — substrates, engineered layers, membranes — is designed to be flexible, and the company is not trying to build end products. It is selling the material, and letting customers build the devices.

From growth to specification

Kara's core claim is that it grows diamond in its own reactors, then processes and characterizes it in-house. That vertical integration is important for two reasons. First, it gives Kara control over the entire supply chain, from growth to final surface finish, which is critical for achieving the sub-nanometer roughness and specified impurity profiles that device makers need. Second, it allows Kara to offer custom fabrication — microfabrication on Kara materials — which is a service that most diamond suppliers do not provide.

The website lists five material categories: substrates (single-crystal wafers in research, optical, and quantum grades), engineered layers (delta-doped and isotopically engineered layers), membranes (thin-film diamond for photonic and quantum devices), thermal materials (heat spreaders and submounts), and custom fabrication. Each category is a different product line, but they all share a common thread: they are specified, grown, and characterized.

That phrase — "Specified. Grown. Characterized." — is the heart of Kara's approach. It suggests that the company is not just growing diamond; it is growing diamond to a customer's exact requirements. That is a very different business model from a typical lab-grown diamond producer, which sells standard sizes and grades. Kara is offering a foundry-like service, where the customer starts with a specification and Kara grows the material to match.

The spec sheet as a sales pitch

Kara's website is unusual in that it is almost entirely a technical document. There is no glossy video, no founder story, no hype. Instead, there is a table of material properties with citations to academic sources (Wort & Balmer 2008, Isberg et al. 2002, Tsao et al. 2018), a chart normalized to 100%, and a note that specification sheets are available under NDA. That is a deliberate choice: Kara is selling to engineers, and engineers trust datasheets more than marketing.

The spec-sheet-first approach also signals a maturity level that is rare in the diamond industry. Most diamond startups are still trying to prove that they can grow large, pure crystals. Kara is implying that the growth problem is solved enough that the company can now focus on precision — on orientation, miscut, surface roughness, impurity profile, and isotope fraction. That is a bold claim, and it is backed by the company's team, which includes people from UC Berkeley, Berkeley Lab, Genentech, Regeneron, Varda Space Industries, and other institutions.

But the spec sheet is also a sales tool. By publishing a detailed comparison table, Kara is making the case that diamond is not a exotic material but a practical one — a material that can be specified, ordered, and integrated into a design. That is the first step toward making diamond an engineering material, not just a scientific curiosity.

The long game: quantum and power electronics

The near-term revenue for Kara is likely to come from thermal management and research substrates. But the long-term opportunity is in quantum and power electronics, where diamond's unique properties could enable entirely new classes of devices.

In quantum sensing, diamond's nitrogen-vacancy centers can detect magnetic fields at room temperature, which could be used for medical imaging, navigation, or materials analysis. In quantum computing, diamond is a leading candidate for spin-photon interfaces, which require precise isotopic control — exactly what Kara's engineered layers promise. In power electronics, diamond's high breakdown field and thermal conductivity could enable switches that operate at higher voltages and temperatures than silicon carbide, with lower losses.

None of these applications is ready for mass production, and Kara is not pretending otherwise. The company's website lists them as "applications" but does not claim to have shipped a quantum computer or a diamond power transistor. Instead, Kara is positioning itself as the material supplier for the people who are building those devices. That is a smart position: it lets Kara focus on what it does best — growing and engineering diamond — while leaving the device design to its customers.

The risk is that the timeline for these applications is long, and the market may not be ready for diamond wafers at the price Kara needs to charge. But if Kara can establish itself as the reliable source for specified diamond, it could become the TSMC of the diamond world — a foundry that enables a new generation of devices.

That is the bet. And it is a bet worth watching.