> Notably, Apple still has not released an Apple silicon chip that can do virtualization
Even if Apple's ARM chips are not classically virtualizable (and I don't know whether that is the case), that was indeed the case for the x86 (pre VT-x) when two fairly straightforward solutions were applied: binary translation of non-virtualizable instructions (VMware approach) and paravirtualization (Xen approach.) This sparked the x86 virtualization revolution even before hardware support was available from intel and AMD.
Other approaches that could potentially work on a "non-virtualizable" ARM CPU include user-mode Linux, emulation, microkernels (similar to paravirtualization - recall Xnu contains mach and presumably still has the ability to outsource system calls and paging) and containers.
Even if Apple's ARM chips are not classically virtualizable (and I don't know whether that is the case), that was indeed the case for the x86 (pre VT-x) when two fairly straightforward solutions were applied: binary translation of non-virtualizable instructions (VMware approach) and paravirtualization (Xen approach.) This sparked the x86 virtualization revolution even before hardware support was available from intel and AMD.
Other approaches that could potentially work on a "non-virtualizable" ARM CPU include user-mode Linux, emulation, microkernels (similar to paravirtualization - recall Xnu contains mach and presumably still has the ability to outsource system calls and paging) and containers.