Re:Huawei builds up 2-year reserve of 'most important' US chips
by etrautmann on Saturday May 30, 2020 @08:13AM
(#23360592)
Yes, these machines are incredibly complex. Even the laser source is a $200M machine from Cymer that implements the same principles as the nations ignition facility by lasing microscopic droplets of molten tin that are falling in a vacuum. That requires rapid closed loop tracking of the tin droplets. It’s perhaps replicable but there’s a lot of art in EUV that took 30 years to develop. It’s not a simple job to clone that.
Re:Huawei builds up 2-year reserve of 'most important' US chips
by ashtonkem on Saturday May 30, 2020 @01:35PM
(#23362942)
And by the time you’ve replicated that, then state of the art will have moved significantly.
Re:Huawei builds up 2-year reserve of 'most important' US chips
by thePunisher on Saturday May 30, 2020 @05:32PM
(#23364952)
It took ASML a long time to perfect these machines, but once they nailed it I'm pretty sure it's no big deal for anyone to duplicate it if they knew how to avoid all the pitfalls.
It would save the Chinese decades if they stole ASML's secrets instead of trying to figure it out on their own. Most likely they will simply take the shortcut and then start selling the machines at half the price ASML is selling them for (as they do with almost everything).
I don't think you can just replicate something like a EUV machine from them in any way.
You probably need highly specialized workshops for tons of different components and all of those workshops have to go through the same innovation cycle.