Good luck AMAT and LAMRevenues of US semiconductor companies in China, in million USD. These includes sales to all customers inside China, both domestic and international companies.
The names of the first six companies in Chinese in the table below are: Qualcomm, Intel, TI, Broadcom, Nvidia and Micron.
View attachment 99643
Good luck all this company after 2024Doanh thu của các công ty bán dẫn của Mỹ tại Trung Quốc, tính bằng triệu USD. Chúng bao gồm bán hàng cho tất cả các khách hàng bên trong Trung Quốc, cả các công ty trong nước và quốc tế.
Tên của sáu công ty đầu tiên bằng tiếng Trung trong bảng dưới đây là: Qualcomm, Intel, TI, Broadcom, Nvidia và Micron.
View attachment 99643
ICRD 一条 国产 线 ,。 涉及 到 有 欧美 的 线 , 就有 一样 的 机器 对 标On SMIC and 14nm/N+1 production, please go read back on my post here
#18,363
SMIC's SN1/SN2 fab cannot use American product, because sub 14nm tools were banned by previous sanctions. According to @PopularScience, that's why expanding production there has been so slow.
From the Lam interview
Similarly, ICRD already has a fully domestic production line and I think it's probably 14 nm or better.
So IMO, N+1 production at SN1 fab is a given at this point. I don't know for sure about N+2, but I don't really see anything that would stop N+2 production. Maybe EDA tools? But if previous roadmap is to be believed, they already have successful produced N+2 chips at SMSC fabs. Maybe the yield aren't very high, but there should be no reason why they can't start mass production sometimes next year.
Again, Chinese fabs do not need to be 100% domestic. They can continue to buy lithography machines from ASML and tools from Japan/SK/Taiwan.
Then, for now, Chinese companies will have to use the more expensive equipment that requires vacuum handling etc until Chinese or non-American companies willing to sell to China can replicate or do better than what LAM has.Everything else is OK but KLA. Advantest cannot replace KLA. Unfortunately KLA has a near monopoly on optical scatterometry for semiconductor QC and the only other company making it is Onto Innovation which is also US based. Everyone else (Chinese companies and Japanese companies like Hitachi) use CD-SEM which is much more expensive and requires vacuum handling i.e. loadlocks, pumps, vacuum grade materials, which imposes high cost and lowers throughput. You get much more detailed information from CD-SEM but that info is often unnecessary for QC alone.
The aim should be for 100% domestic. Foreign countries cannot be trusted to remain commercially rational as far as the interests of their companies is concerned. If there is one country that can afford to have the entirety of the supply chain for manufacturing equipment and operations located domestically, it is China. Eventually, this should be the goal.On SMIC and 14nm/N+1 production, please go read back on my post here
#18,363
SMIC's SN1/SN2 fab cannot use American product, because sub 14nm tools were banned by previous sanctions. According to @PopularScience, that's why expanding production there has been so slow.
From the Lam interview
Similarly, ICRD already has a fully domestic production line and I think it's probably 14 nm or better.
So IMO, N+1 production at SN1 fab is a given at this point. I don't know for sure about N+2, but I don't really see anything that would stop N+2 production. Maybe EDA tools? But if previous roadmap is to be believed, they already have successful produced N+2 chips at SMSC fabs. Maybe the yield aren't very high, but there should be no reason why they can't start mass production sometimes next year.
Again, Chinese fabs do not need to be 100% domestic. They can continue to buy lithography machines from ASML and tools from Japan/SK/Taiwan.