Its hard, but the value of the software is a massive effort, that cannot be translated or reproduced under a new ISA.
I think we both agree that in x86 should have been retired a long time ago, but its only alive because of the legacy that accompanies it.
> The hard part is ecosystem adoption, drivers, compilers, existing low-level code written for the ISA.
Exactly. There are so many parts and probably many more. The volume of effort to complete that for a new platform will probably not be done. Apple is a different case because they control the MAC platform in full.
> I have no doubt Huawei can create a different ISA if they want to.
Ive also played tennis with people who said they can win Wimbledon, and until THEY win, i dont believe them.
The only way to know for sure is to have a new ISA, but H have yet to try.