Leo from Kitguru made an interesting point on moore’s law show this week, that this perpetual state of “might be forced to cancel” and “products might be banned” actually probably serves to funnel inventory to china even faster. like now it’s a race against time to get the inventory there before whatever sanction this week, and you better get as much other stuff over there before it gets banned too.
The sanctions are generally ineffective, china already pushed up to 7nm DUV far quicker than anyone expected, and if they develop indigenous EUV then the silicon shield is over. I know everyone thinks that’s impossible for anyone to catch up to ASML/etc, but china doesn’t have to respect western IP either, it’s impossible to catch up to ASML under the IP regimes set up by western governments but it’s significantly easier for them. “Can it be done” is a question of willpower and spending and time, yes it will be extraordinarily difficult (replicate the quality of zeiss optics, ASML EUV sources, etc) but they’ll get there eventually, and probably far faster than anyone on the outside assumes, just like with 7nm tier nodes. It’s a lot easier to build the second fusion reactor when you know how the first one works, that whole “fast follower” thing.
And when that happens you will have a country completely unencumbered by any of the patent-system deadweight loss, or the other problems of late-stage capitalism. It is like Iran: not only did the sanctions not change anything, but now you have a country that has been forced to develop indigenous industry and technology, and they’re still just as bad as ever but now they’re bad and also have a stream of technology that you don’t control.
Leo from Kitguru made an interesting point on moore’s law show this week, that this perpetual state of “might be forced to cancel” and “products might be banned” actually probably serves to funnel inventory to china even faster. like now it’s a race against time to get the inventory there before whatever sanction this week, and you better get as much other stuff over there before it gets banned too.
The sanctions are generally ineffective, china already pushed up to 7nm DUV far quicker than anyone expected, and if they develop indigenous EUV then the silicon shield is over. I know everyone thinks that’s impossible for anyone to catch up to ASML/etc, but china doesn’t have to respect western IP either, it’s impossible to catch up to ASML under the IP regimes set up by western governments but it’s significantly easier for them. “Can it be done” is a question of willpower and spending and time, yes it will be extraordinarily difficult (replicate the quality of zeiss optics, ASML EUV sources, etc) but they’ll get there eventually, and probably far faster than anyone on the outside assumes, just like with 7nm tier nodes. It’s a lot easier to build the second fusion reactor when you know how the first one works, that whole “fast follower” thing.
And when that happens you will have a country completely unencumbered by any of the patent-system deadweight loss, or the other problems of late-stage capitalism. It is like Iran: not only did the sanctions not change anything, but now you have a country that has been forced to develop indigenous industry and technology, and they’re still just as bad as ever but now they’re bad and also have a stream of technology that you don’t control.