This is exactly what I want to research more and write (glad I found out before going into the weeds). I really like the perspectives and totally agree that the US and China have shared problems with data centers. That being said, I think location is not the only reason. The broader environmentalist movement/concern in China is not as big as in the US, and people do not have much say over what lands can be used for what (as most lands are owned by the state). I also think that Chinese society is much more acceptable about big construction work.
Thank you for your comment! That said, I think bringing up ASML and reverse engineering may be slightly off-topic here, as this piece is really more about tech industry governance rather than a straightforward "who has the better chips" race.
Based on current reporting from mainland China, the focus seems to be less on building a single, extremely precise lithography machine, and more on integrating the entire lithography workflow end-to-end. The expected window for showcasing results from this "workflow integration" approach is generally seen as 2028 to 2030.
This is exactly what I want to research more and write (glad I found out before going into the weeds). I really like the perspectives and totally agree that the US and China have shared problems with data centers. That being said, I think location is not the only reason. The broader environmentalist movement/concern in China is not as big as in the US, and people do not have much say over what lands can be used for what (as most lands are owned by the state). I also think that Chinese society is much more acceptable about big construction work.
China will be able to reverse engineer ASML technology by 2031. Until China has compute parity, it's not a real race.
First inning two outs is 2026. The real race starts about 2035.
Thank you for your comment! That said, I think bringing up ASML and reverse engineering may be slightly off-topic here, as this piece is really more about tech industry governance rather than a straightforward "who has the better chips" race.
Based on current reporting from mainland China, the focus seems to be less on building a single, extremely precise lithography machine, and more on integrating the entire lithography workflow end-to-end. The expected window for showcasing results from this "workflow integration" approach is generally seen as 2028 to 2030.