i regret what i said in the past

@_realnickk Wut? It's called a virtual machine, and without hardware virtualization, it is almost impossible to properly utilize it. Most Raspberry Pis can barely handle most processor intensive computation tasks in the first place, 'emulating' x86_64, even through single-core virtual machines would make everything mostly unusable.

Everything can technically emulate everything, in most cases, the question is what limits you personally have, and whether the software runs acceptibly, in the case of a Raspberry Pi, those bounds are pushed to places where most individuals wouldn't want to touch, and for a reason. This is like getting Counter-Strike running on a smart watch, doesn't make much sense.

Just to add, you wouldn't even get any kind of virtualization / virtual machine environment without having GNU + Linux as your bare metal operating system, which mostly defeats the point of having a secondary operating system...

@_realnickk I was mainly talking about virtual machines because at this stage, emulating x86 software in ARM is a joke.

Talking objectively, Windows 10 ARM is much slower than a regular Windows 10 release. Emulating x86 in ARM, or practically emulating any instruction set causes quite big performance penalties. This has been tested on release, mainly by Techspot, which did some really great benchmarks.

Candidates:

1) i7-8550U - Laptop processor, natively on x86_64 (not emulating) - quad(4) core, hyperthreaded(8), up to 4 GHz boost clock.

2) i5-5200U - Laptop processor, natively on x86_64 (not emulating) - dual(2) core, hyperthreaded(4), up to 2.7 GHz boost clock.

3) Celeron N3450 - Laptop processor, natively on x86_64 (not emulating) - quad(4) core - non-hyperthreaded(4), up to 2.2 GHz boost clock.

4) Snapdragon 835 - Mobile processor, natively on ARM (emulating) - octa(8) core - non-hyperthreaded (clustered) up to 2.45 GHz boost clock.

On Cinebench R11.5, the results are pretty self-explanatory, both the singlecore ones and multicore ones...

1) SC - 1.8, MC - 7.0

2) SC - 1.3, MC - 2.9

3) SC - 0.5, MC - 1.7

4) SC - 0.3, MC - 1.6

Leading to the simple conclusion that an obviously faster performing and newer processor simply performs worse in a workload emulated from x86 to ARM. I get your point that it's **possible** to do it, but it's not practical outside of any reason other than that there's technology to do it. QEMU-KVM would probably suffice for the virtualization requirements, as Virtualbox is often unused in Linux due to it being bloated.