AFAIK 28nm isn't that hard to reach - the cutting edge is like, 5nm/3nm right now? SMIC certainly would face demand for 28nm chips - there's a very strong appetite within China for indigenization of these technologies, because a major consequence of the Trump admin's crackdown with export controls and everything was to increase appetite from Chinese companies for domestically produced products that were formerly viewed as inferior. Dan Wang has a good article on this: https://www.foreignaffairs.com/articles/united-states/2021-07-29/chinas-sputnik-moment
That said, ramping up chip production at smaller nodes is a nontrivial task. Huge capital outlays, a lot of tacit knowledge required, and then the other difficult-to-acquire technological inputs like EDA tools and lithography machines. So not pretending like this will happen overnight or this is a sure thing, and my confidence intervals are fairly large to reflect that. For SMIC, things could pretty plausibly go either way.
With that said, I think there's room for optimism. Two reasons: (1) within China, gov't, manufacturers, and consumers are all swimming in the same direction (more appetite for domestically produced chips), and I think that means the domestic chip manufacturing ecosystem is more likely to see success; (2) it will on the whole get easier and easier to make 28nm chips anyway, because of the general advancement of technology. So estimating a modestly accelerating upward trend.
AFAIK 28nm isn't that hard to reach - the cutting edge is like, 5nm/3nm right now? SMIC certainly would face demand for 28nm chips - there's a very strong appetite within China for indigenization of these technologies, because a major consequence of the Trump admin's crackdown with export controls and everything was to increase appetite from Chinese companies for domestically produced products that were formerly viewed as inferior. Dan Wang has a good article on this: https://www.foreignaffairs.com/articles/united-states/2021-07-29/chinas-sputnik-moment
That said, ramping up chip production at smaller nodes is a nontrivial task. Huge capital outlays, a lot of tacit knowledge required, and then the other difficult-to-acquire technological inputs like EDA tools and lithography machines. So not pretending like this will happen overnight or this is a sure thing, and my confidence intervals are fairly large to reflect that. For SMIC, things could pretty plausibly go either way.
With that said, I think there's room for optimism. Two reasons: (1) within China, gov't, manufacturers, and consumers are all swimming in the same direction (more appetite for domestically produced chips), and I think that means the domestic chip manufacturing ecosystem is more likely to see success; (2) it will on the whole get easier and easier to make 28nm chips anyway, because of the general advancement of technology. So estimating a modestly accelerating upward trend.