> Has it been demonstrated that RISC-V is architecturally suitable for making chips that equal the performance of high-end x86 and ARM designs?

How would you demonstrate that besides actually building such a chip?