Somewhere inside a cleanroom you will never enter, a machine the size of a small bus is hurling 13.5-nanometer photons at a sheet of silicon. A few feet away, bolted to a vacuum chamber, sits a smaller instrument that was almost certainly built in Martinez, California. It is measuring what those photons just did. Without that smaller instrument, nobody on the production line would be sure the bigger one was working.
That smaller instrument is the business of EUV Tech. The company does not make chips. It does not make scanners. It makes the rulers, the witnesses, the second opinions - the at-wavelength metrology tools that let extreme ultraviolet lithography stop being a physics demo and start being a manufacturing process. There are perhaps a dozen places on earth where this matters. EUV Tech is in all of them.
The chip industry spent two decades and tens of billions of dollars learning to print with 13.5 nm light. Then it needed someone to check the work.
- The premise of EUV Tech, distilledA logo, registered. Most people will never see one of EUV Tech's instruments. Their phones, laptops and cars depend on them anyway.
// 01 - THE PROBLEMWhat 13.5 nanometers does to your assumptions
Extreme ultraviolet lithography is the technology that keeps Moore's Law on life support. It uses light with a wavelength of 13.5 nanometers - roughly fifty times shorter than the visible photons coming out of your screen right now. At that wavelength, almost every material that used to be transparent becomes opaque. Glass lenses stop working. Air stops working. You have to run the whole optical path in a vacuum, bouncing the light off mirrors coated with dozens of alternating atomic layers. It is, in every literal sense, rocket science applied to printing.
Which raises an awkward question. If your lithography light is so exotic that ordinary optics cannot handle it, how do you measure anything about your masks, your pellicles, your mirrors and your resists before you commit them to a $200 million scanner? You can fake it with deep ultraviolet probes. You can guess with electron beams. Or you can do the thing that sounds obvious and is actually fiendish: build a small, dedicated instrument that uses the same wavelength of light the scanner uses, and measure with that.
EUV Tech does the third one. Almost nobody else commercially does.
You can model what 13.5 nm light will do to your mask. Or you can shine 13.5 nm light at your mask and look. EUV Tech sells the second option.
- Why at-wavelength metrology exists// 02 - THE FOUNDERS' BETA Berkeley Lab spinoff with a 25-year fuse
EUV Tech was incorporated in 1997. To put that in context: production EUV lithography did not exist. ASML's first commercial EUV scanner would not ship for another two decades. The company was founded the same year a young researcher named Patrick Naulleau joined Lawrence Berkeley National Laboratory, where the Center for X-Ray Optics - CXRO - was quietly building the soft x-ray and EUV science that would, eventually, become a global industry.
The bet was simple, if not exactly safe: build the at-wavelength measurement infrastructure the EUV industry would need, and wait. For most of the company's existence, the wait was the hard part. Then EUV went into production. Then Moore's Law refused to die. Then everyone in the supply chain suddenly needed exactly what EUV Tech had been quietly perfecting in a low-slung industrial building near the Carquinez Strait.
In late 2022, Patrick Naulleau left CXRO - he had become its director in 2010 - to take over as CEO of EUV Tech. The man who had spent 25 years building the science was now running the company built to commercialize it. A few months later, Intel Capital led the Series A.
The CEO, in one paragraph
Patrick Naulleau spent a quarter century at Berkeley Lab working on EUV lithography, nano-diffractive optics and soft x-ray microscopy. He ran CXRO for over a decade. He has more peer-reviewed papers on EUV than most companies have employees. In October 2022 he became CEO of EUV Tech. The transition felt less like a career change than a logical conclusion.
// 03 - THE PRODUCTThe catalog nobody outside the industry has ever read
EUV Tech sells instruments with names that look like Scrabble accidents. Reflectometers. Pellicle deflection tools. N&K and phase measurement systems. Actinic mask defect review platforms. Zoneplate microscopes. Each one solves a specific problem that you only have if you are trying to push 13.5 nm light through a manufacturing process at meaningful yield.
The hardware
The reflectometers measure how much EUV light bounces off a multilayer mirror or mask blank, at exactly the angles and wavelengths the scanner will use. The pellicle tools measure the gossamer thin membranes that protect EUV photomasks from particles - membranes that have to be thin enough to be nearly invisible to 13.5 nm light, strong enough not to vibrate themselves apart, and uniform enough not to print as a defect. The N&K tool, delivered to customers in late 2022, measures the optical constants of new EUV materials so the resist chemists and mask makers can stop guessing.
The software
SuMMIT - the company's image analysis suite, sometimes branded as Lithometrix - is the de-facto industry tool for measuring line-edge roughness and critical dimensions out of scanning electron microscope images. It is one of those quiet pieces of software that ends up cited in nearly every academic EUV paper and installed on nearly every fab metrology workstation, without anyone making a fuss about it.
The services
If you cannot afford to install a half-million-dollar at-wavelength instrument of your own - and you probably cannot - EUV Tech will measure your samples for you. They run their own tools in their own building. It is the kind of pragmatic side business that only exists when the hardware is genuinely scarce.
You don't buy an EUV reflectometer because you want one. You buy it because the alternative is shipping a $20 million mask and hoping.
- The unspoken sales pitchMilestones, in chronological order
A timeline. The years on the left look ordinary. The decisions behind them were not.
// 04 - THE PROOFWhat the numbers will and will not tell you
EUV Tech is a private company, and a discreet one. The Series A amount was never disclosed. Reported annual revenue hovers around four million dollars - which would be unremarkable for a software company and is roughly the price tag of a single high-end tool. What the revenue figure undercounts is concentration: a handful of customers, each of them a household name in semiconductors, each of them returning year after year.
Where the work goes
Bars indicate relative emphasis across EUV Tech's public-facing business lines. Hardware dominates; software and services punch above their headcount.
A four-million-dollar revenue line item, sitting in front of a trillion-dollar industry. That is a lever, not a footnote.
- The math, charitably interpreted// 05 - THE MISSIONLighting the way, with apologies for the pun
The official tagline is "lighting the way to a brighter future." Yes, it is a pun about photons. The company is allowed one. The deeper version is this: EUV lithography is the bottleneck and the breakthrough of advanced chipmaking, and the only way to keep pushing it is to measure it honestly. EUV Tech exists to make that honest measurement possible.
That has translated, in practice, into a culture that is heavier on PhDs than on marketers. The website still runs on Squarespace. The press releases are sparse and informational. The growth in headcount - from roughly 70 employees before the Series A to about 110 now - has gone into engineers and applications scientists, not sales theater.
Customers, partnerships, alibi
- Intel Capital led the 2023 Series A. The corporate parent has obvious reasons to want this supplier strong and independent.
- Lawrence Berkeley National Laboratory / CXRO is the research lineage. The CEO ran it. The technology grew up there.
- Global EUV supply chain - foundries, IDMs, mask blank suppliers, pellicle makers, national labs - shows up as customers in product announcements and conference papers, by inference rather than by named logo.
// 06 - WHY IT MATTERS TOMORROWHigh-NA, and the next wavelength of problems
The next generation of EUV scanners - High-NA EUV - is rolling into fabs now. The optics are bigger. The masks behave differently. The pellicles get harder. Every parameter that EUV Tech already measures becomes more sensitive, and a few new ones get added to the list. The instruments will need updating, and the company is openly building toward that horizon.
Underneath High-NA sits a longer question. The semiconductor industry has bet roughly $30 billion of capital expenditure per leading-edge fab on the assumption that EUV will keep delivering. That bet only pays out if the metrology keeps up. If at-wavelength measurement falls behind, yields fall behind, and the economics of the entire roadmap wobble.
So the small instrument is still bolted to the vacuum chamber, a few feet from the bus-sized scanner, in a cleanroom you will never enter. It is still measuring what the 13.5 nanometer photons just did. The scanner gets bigger, the masks get trickier, the pellicles get thinner - and the small instrument keeps quietly upgrading itself to keep up.
That is EUV Tech's job. It has been EUV Tech's job for twenty-eight years. It will probably be EUV Tech's job for the next twenty-eight.
The bus-sized scanner gets the press release. The small instrument next to it gets the truth.
- The shape of the bet, unchanged since 1997// 07 - WHERE TO GO NEXTLinks, news, and the company itself
A directory. The press releases are short, the product PDFs are dense, the talks are worth the hour.
Contact
EUV Tech, Inc.
2830 Howe Road, Suite A, Martinez, CA 94553
+1 (925) 229-4388
info@euvtech.com · pnaulleau@euvtech.com