In letters to the Lower House of the Dutch parliament, trade minister Sjoerdsma said the government of the Netherlands opposes the extraterritoriality of the MATCH Act, since it believes countries should be responsible for their own export controls. Apparently, Dutch diplomats and visitors have been pushing this message in DC recently, including to Congress:
The irony writes itself: Congress wants to bind ASML through extraterritorial reach while the executive that nominally runs controls has not substantively updated them since December 2024 and spent the past year arguing over whether to sell H20s and H200s at all. Sjoerdsma's "own your own controls" line lands harder when Washington can't decide what its own policy is. Thanks for the Tweede Kamer links, that primary source thread is hard to find from this side of the Atlantic.
I think this arguement is less about chips than it is about technological ecosystems.
Machines can be purchased. Designs can be copied. Parts can often be reverse engineered. But some forms of knowledge are much harder to duplicate. Maintenance expertise, supplier relationships, process knowledge, engineering talent, and institutional experience accumulate over decades.
That is what makes the servicing question so fascinating. The debate is not simply whether China can buy advanced equipment. It is whether access to the broader ecosystem surrounding that equipment can be restricted long enough to preserve a technological lead.
At the same time, history suggests that challengers eventually develop many of these capabilities themselves. The question is not whether that happens, but how long it takes and what strategic advantages the delay creates for the incumbent.
For me, that is what makes this such an interesting industrial policy story. It sits at the intersection of technology, manufacturing, institutions, and geopolitics.
the cxmt inclusion is the one worth watching. companies don't just 'use' dram, they design it in. mid-qualification cycles are 12-18 months. locking that in via legislation means any chinese fab that gets qualified for a product before the bill passes becomes a permanent supply-chain problem for that product, not a procurement switch you can flip. that's a different kind of control than the equipment side.
Some more background:
In letters to the Lower House of the Dutch parliament, trade minister Sjoerdsma said the government of the Netherlands opposes the extraterritoriality of the MATCH Act, since it believes countries should be responsible for their own export controls. Apparently, Dutch diplomats and visitors have been pushing this message in DC recently, including to Congress:
https://www.scmp.com/news/china/diplomacy/article/3353452/netherlands-protests-us-proposal-further-bar-chip-giant-asml-china-market
Dutch letters:
https://www.tweedekamer.nl/kamerstukken/kamervragen/detail?id=2026Z07615&did=2026D21750
https://www.tweedekamer.nl/kamerstukken/kamervragen/detail?id=2026Z10558&did=2026D29627
The irony writes itself: Congress wants to bind ASML through extraterritorial reach while the executive that nominally runs controls has not substantively updated them since December 2024 and spent the past year arguing over whether to sell H20s and H200s at all. Sjoerdsma's "own your own controls" line lands harder when Washington can't decide what its own policy is. Thanks for the Tweede Kamer links, that primary source thread is hard to find from this side of the Atlantic.
I think this arguement is less about chips than it is about technological ecosystems.
Machines can be purchased. Designs can be copied. Parts can often be reverse engineered. But some forms of knowledge are much harder to duplicate. Maintenance expertise, supplier relationships, process knowledge, engineering talent, and institutional experience accumulate over decades.
That is what makes the servicing question so fascinating. The debate is not simply whether China can buy advanced equipment. It is whether access to the broader ecosystem surrounding that equipment can be restricted long enough to preserve a technological lead.
At the same time, history suggests that challengers eventually develop many of these capabilities themselves. The question is not whether that happens, but how long it takes and what strategic advantages the delay creates for the incumbent.
For me, that is what makes this such an interesting industrial policy story. It sits at the intersection of technology, manufacturing, institutions, and geopolitics.
the cxmt inclusion is the one worth watching. companies don't just 'use' dram, they design it in. mid-qualification cycles are 12-18 months. locking that in via legislation means any chinese fab that gets qualified for a product before the bill passes becomes a permanent supply-chain problem for that product, not a procurement switch you can flip. that's a different kind of control than the equipment side.