Quote:

You mean, really unusably slowly?

PPC is the worst case scenario to emulate: multiple instruction sets, FPU and Altivec, MMU, special purpose registers, condition registers, more than one instruction set, decrementer, etc etc.

And x64 is the worst case scenario to emulate it: all that would need to be done on software, with no hardware acceleration on a single thread on a CPU family that now routinely requires 16 or more cores to meet its true speed, and which is architecturally and logically very different.

I'd say it might be decades before something like QEMU was fast enough, but it's already been decades, and the core increase seems to be pushing us further rather than closer, so I'd actually be quite comfortable predicting it never could.