I'm talking specifically about comparatively new process tech/smaller gate widths and I'm hoping the answer isn't as simple as, "it's expensive" or "export restrictions".
That is to say, is the challenge technical (if I was a rich supervillain, could I hire a team to build my own ASML equivalent - or is that form of lithography so complex that it's hard to replicate ). Say I could buy the machines, is it then only a matter of money to set up the fab to production levels ? Or is that a technical gamble too ?
I see a lot of discussion about other countries trying to get into the fab business and I guess, with my lack of understanding of the technical challenge, I always wonder why a state-level entity can't just pay X billion dollars for what is, arguably, an amazingly valuable asset.
I've heard (apocryphal?) stories, here on HN and elsewhere about countries trying (at least on paper) and failing - a recent attempt in China (failed for several reasons), past attempts in India (failed on infrastructure requirements). It was pretty fascinating from a bird's-eye view of the supply chain that results in a chip today.