Then if I understood well what you said: the main bottleneck for many of those RISC-V microarchitectures is the access to the best silicon process. (ofc, they will have to optimize them for this node process)