alt Hacker NewsThis bill essentially creates a legal basis for the U.S. to forbid its companies from servicing semiconductor tool makers if those tool makers do not fit themselves with a proper yokel in 150 days[^1].
In practical terms, this bill is the equivalent of the major of a village forbidding the local blacksmith from making hammers for the goldsmith living in the next village, if said goldsmith sells jewelry to the vast enclave of dwarfs living under the mountain range.
On the enforceability front though, I believe that ASML uses enough American parts and services to be forced into compliance at least for half a decade, though I wish they would start unentangling from any American dependencies immediately.
[^1]: Page 12, lines 22-24
Replies
It is unfortunate that the people in power are too fucking mentally incapable to comprehend even the most basic of game theory. The US has no leverage over you unless you surrender pre-emptively. The correct move here is to call their bluff - if the US wants to cut off supply to ASML, ASML cuts off supply to the US. The US cannot live without ASML anymore than ASML can live without it, so they would be forced to back down. Yet instead we're in a trivial prisoner's dilemma where you get the second move with full information - you know your opponent has picked to defect - and you still choose cooperate, even though it's strictly incorrect. Led by donkeys.
It is all about ASML and preventing their business with China.
Incredible what the EU puts up with.
I suspect the intent (hope) is that by then there will be more fabs running in the US and that we won't have cut our legs off at the knees. It's pretty hard to see a significant chunk of chip manufacturing being onshored by then though, if it ever happens.
ASML licenses EUV technology from the US government, which developed it. It is physically manufactured in the US, possibly as a condition of the license agreement. This is where the leverage of the US government comes from. To disentangle, ASML would have to develop an independent EUV technology that is practically substitutable. They have an existing installed base they need to continue to maintain.
They may be able to do this but it would likely require many, many years before they could sunset their current EUV license. It could make more sense to just work on whatever will eventually replace EUV.