So following up on this point though, I actually just had a conversation with some coworkers about simulation software last week, and apparently you have to do a lot of discrete modeling with experimental validation once you get to more complex dynamics. I wouldn’t be surprised if a lot of the technical insufficiencies of Chinese solutions come down to extensiveness of discrete modeling libraries. Those are the sorts of things you need to build up experimentally and would be very valuable to development work, enough to want to discriminate against different software packages.the article is paywalled, so I will just respond based on your comment and other's comments.
imo, there is nothing wrong with the Chinese alternative here. Writing simulation software isn't difficult. It's possible that American chips and software are a little better, so Chinese MIC used them. But if they are not available, Chinese MIC will just use domestic alternatives. People that write these articles fundamentally don't seem to understand anything about technology.