This is the reason why chip manufacturing is not just about buying an EUV machine and then starting to pump out chips to make it rain, as many people believe.
This is craftsmanship. John Doe can use Python to create software, and so can you, yet John Doe's code runs better and faster while your code crashes all the time.
People seem to forget that a craftsman's ability to use tools is a big factor in the final product.
Many people are saying it's all about the Zeiss machines and TSMC isn't actually bringing much over that. Usually as an argument about EU tech competitiveness.
Both. There's a huge difference between designing a functional chip and designing a chip with high yield and long term reliability. Companies and individuals vary wildly in their processes and understanding.
It's becoming a wide spread problem in the industry. For decades technology-minded folks have been told to go into software development because that's where the money is. The field of custom IC design has been both short on applicants and has had huge barriers to entry. As a result, the experienced engineers are retiring and there aren't enough juniors coming up to replace them. Skills and institutional knowledge are being lost.
There is a lot of focus right now on tools such as AI to allow junior people to produce things at an expert level based on encoding that expert knowledge. It would be great if it worked, but so far the results aren't there.
Source: I develop EDA software for custom IC design.
This is craftsmanship. John Doe can use Python to create software, and so can you, yet John Doe's code runs better and faster while your code crashes all the time.
People seem to forget that a craftsman's ability to use tools is a big factor in the final product.