mmTron.

There is a rule in radio engineering that everyone learns and no one likes: the faster you go, the harder it is to stay honest. Push a signal up into the tens of gigahertz, ask an amplifier to make it louder, and the amplifier starts to lie a little - it clips, it distorts, it smears the clean waveform you sent into something the receiver has to squint at. Fixing that is unglamorous, expensive, and absolutely necessary if you want satellite internet, 5G that actually works, or a radar that can tell a drone from a bird. mmTron, a fabless semiconductor company in Redwood City with fewer people than a mid-size restaurant, exists to fix exactly that.

The name is a tell. "mm" is for millimeter-wave, the band of spectrum from roughly 30 to 300 gigahertz where the wavelengths are measured in millimeters and the engineering gets genuinely hard. This is where wireless goes when it runs out of room lower down. There is a lot of bandwidth up there, which is why everyone - satellite constellations, 5G and 6G planners, defense programs, chipmakers building test gear - wants to use it. There is also a lot of physics up there that makes components misbehave, which is why relatively few companies can build parts that work well. mmTron is betting the whole company on being one of them.

What it actually makes are MMICs - monolithic microwave integrated circuits - and the closely related RFICs. In plain terms: tiny chips that amplify, mix, switch and generate very high-frequency signals. The catalog is broad for a company this size: high-linearity power amplifiers, low-noise amplifiers, front-end modules, mixers, frequency multipliers, voltage-controlled oscillators, switches, and wideband distributed amplifiers that deliver gain from near-DC to past 100 GHz on a single part. If a system runs above about 30 GHz, mmTron would like to be somewhere inside it.

That list of materials is not marketing filler - it is the strategy. Most small chip companies pick one process technology and marry it, because supporting more than one is expensive. mmTron deliberately works across five: gallium arsenide, gallium nitride, indium phosphide, silicon-germanium and RF silicon-on-insulator. Each has a sweet spot. GaN gives you output power; InP gives you frequency; SiGe and RFSOI give you integration and cost. By refusing to commit to one, mmTron can match each chip to the process that suits it rather than forcing every product through the same door. It is more work. It is also, arguably, the moat.

The founder is Dr. Seyed A. Tabatabaei, who is President, CEO and chairman, and who did not arrive at this cold. Before mmTron he spent nine years running Teramics, a millimeter-wave design-services house. Nine years of building mmWave circuits for other people is nine years of watching which parts customers kept asking for and couldn't buy off a shelf. mmTron reads, in the most literal sense, like a list of those missing parts turned into a company. He co-founded it in 2020 with Mona Molaasgari, who serves as COO and President. He holds a PhD from the University of Maryland and a management certificate from MIT Sloan; the second credential matters more than it sounds, because turning a design-services instinct into a product company is mostly an operating problem.

The fab you never see

mmTron is fabless, which means it does not own a chip factory - and here the interesting part is how far it takes the idea. Being fabless usually means designing a chip and handing it to a foundry. mmTron instead hand-selects the entire chain itself: which foundry runs the wafers, who does the packaging, how the parts are tested in production, and how they get qualified for reliability. It is design-only on the org chart and end-to-end in practice. For a company selling into satellites and defense radar - places where a part failing in the field is not an option - owning that whole flow without owning a single cleanroom is the point. You get to promise reliability without carrying the capital cost of a fab.

Who buys this? The clearest customer is space. Low-Earth-orbit mega-constellations - the swarms of satellites promising broadband from orbit - live and die on their RF links. Moving high-speed data between satellites and down to Earth demands amplifiers that are both highly linear and power-efficient, because every watt wasted in orbit is a watt you had to launch. mmTron's V-band amplifiers, the TMC315 and TMC316, are aimed squarely at satellite-to-Earth downlinks in the 37-to-43-gigahertz range. The same appetite for clean, high-frequency signal handling shows up in 5G and 6G infrastructure, in defense and commercial radar, in data centers, and in the test-and-measurement instruments that everyone else in this industry uses to check their own work.

The money arrived in April 2026, when mmTron announced a $45 million Series A led by Maverick Silicon, with Tipping Point, J&M Copper Beech Ventures, Silicon Catalyst Ventures and a roster of angel groups alongside. Spread across 18 people, that is roughly $2.5 million raised per employee - the kind of ratio you only see in deep-tech, where investors are paying for scarce design talent rather than a finished market. The stated use of proceeds is unromantic and correct: accelerate development and production of high-linearity mmWave products, and build toward system-level platforms rather than just selling parts one at a time.

The other signal is on the board. mmTron recruited Diane M. Bryant, who spent 35 years in technology leadership and, at Intel, ran the roughly $19 billion Data Center Group before stints at Google Cloud and elsewhere. Executives of that altitude do not join 18-person startups for the meeting cadence. When someone who ran one of the largest infrastructure businesses in the industry takes a board seat at a millimeter-wave chip company, it is worth reading as a view on where the market is going.

What you can actually do with it

It helps to be concrete about what these chips let people build, because "MMIC" is not a word that means much outside the trade. An engineer designing a phased-array antenna for a satellite terminal needs dozens of small amplifiers behind the array, each one nudging the signal without distorting it; mmTron's front-end modules and power amplifiers are the parts that go there. A team building a 5G small cell that operates in the 24-to-40-gigahertz bands needs the same class of components with different priorities. A defense contractor building a W-band radar, or an instrument company building a signal analyzer that has to measure other people's mmWave chips, is shopping in exactly this aisle. In each case the buyer is not an end consumer but another engineer, which is why mmTron sells the way it does - a catalog of standard parts plus custom designs, distributed globally through Richardson RFPD so a design team in another country can order a sample without a sales call.

The other thing worth saying is that mmTron is trying to move up the stack. Selling individual amplifiers is a fine business, but a crowded one. The stated ambition behind the Series A is "system-level solutions" - not just the chip, but the front-end platform the chip lives in. That is a meaningfully harder thing to build and a meaningfully stickier thing to sell, because a customer who designs your platform into their satellite is not going to swap it out casually. It is the difference between selling bricks and selling walls.

None of this makes mmTron a sure thing. It is small, it is early, and it is competing for attention with far larger RF suppliers - Qorvo, Analog Devices, MACOM, Mini-Circuits - that have their own mmWave lines and decades of relationships. The mmWave market is real but still emerging; a lot of the biggest applications, from 6G to the largest satellite constellations, are years from full scale. A company burning $45 million has to time its products to a market that is arriving on its own schedule, not mmTron's. What mmTron has going for it is focus, a founder who spent nearly a decade cataloguing exactly which parts the market wanted, a multi-process strategy that lets it say yes to more of them, a board member who has run infrastructure at genuine scale, and now the capital to build faster than a bootstrapped design house ever could. In a business where the winners are usually the ones who picked the right chip for the right job instead of forcing one answer everywhere, that is a coherent bet. The millimeters, it turns out, are worth quite a lot.