America & the west has been way way in the lead with (communist-ic) giving away of chips designs, open source. It'll be interesting to see if CAS really is able & willing to contribute.

One place i think they could really make a mark: there's a lot of un-core that is still pretty hard to replicate, inaccessible. Trying to make more interfaces available, commodity, rather than bitterly guarded IP that a scant few companies own & rent out (that i believe even folks like SiFive simply go shopping for)... that could really change the game, help make chipmaking much more diverse again.

Otoh simply cranking out good solid risc-v core innovation is also needed. It'll be interesting to see whether such designs are flexible, iterable, evolvable, or fixed & monolithic in mature.