Upstream is where the leverage is, and in EUV the leverage runs deeper than the scanner. Everyone knows ASML is the sole maker of EUV lithography systems. Fewer people trace it one step further: the optics inside those systems — the mirrors that steer 13.5nm light — come from a supply chain just as narrow, dominated by Carl Zeiss. A chokepoint inside a chokepoint.
Why mirrors and not lenses? Because EUV light is absorbed by essentially everything, including glass, so you cannot refract it through a lens — you have to bounce it off precision reflective surfaces. Carl Zeiss's grant US12650544B2, "Method for producing an optical element" (issued June 9, 2026; CARL ZEISS SMT GMBH; CPC G02B5/18 multilayer-mirror classes), is about manufacturing those elements. The multilayer coatings on an EUV mirror have to reflect a usable fraction of an already-scarce beam, and producing them to that tolerance is its own discipline.
Then the optics have to stay perfect while absorbing heat, which is where measurement comes in. ASML's grant US12645147B2, "Temperature measurement of optical elements in an optical apparatus" (issued June 2, 2026; ASML Netherlands B.V.; CPC G03F7/70xxx lithography classes), is about knowing the optic's temperature precisely. The reason is physics: even tiny thermal expansion shifts a mirror's figure by enough to blur a sub-2nm pattern. You cannot correct a thermal distortion you cannot measure, so the metrology is inseparable from the optics.
Put the two grants side by side and you see the full demand: build the mirror to near-atomic tolerance, then keep it there under a heat load, in real time. That combination — manufacturing precision plus active thermal management — is why the EUV optics supply chain is measured in a handful of companies, not a market. Substitutability is essentially zero on a meaningful timescale.
The export-controls angle, since this is the controls desk: when policy debates talk about restricting "EUV," the implicit chokepoint is this entire stack — scanner, optics, metrology — not a single box. The narrower the supplier base, the more leverage a control regime has, and the optics layer is among the narrowest links in the whole semiconductor supply chain. Controls reshape the addressable market precisely because there is no second source to route around.
So the headline "ASML monopoly" undersells the concentration. The 2026 record from Zeiss and ASML shows that even within that monopoly, the optics and their thermal metrology are a distinct, equally narrow dependency. Follow the tool, not the chip — and then follow the mirror inside the tool. That is where the real leverage in leading-edge lithography sits.