Two points

Two points:

1) It's surprisingly easy to get the Hurd running in a
virtual machine with kvm:

$ uname -a
GNU hurd 0.9 GNU-Mach 1.8+git20240714-up-amd64/Hurd-0.9 x86_64 GNU

2) I own two Mac mini's, which are sitting in a drawer.  They
were made from notebook equipment, and they are crap.

One of them came with 4GB of RAM, and ran like a dog.  Thankfully,
I could add more (notebook) RAM, bringing it up to 16GB.  Then
it ran acceptably, probably because it was caching the slow, slow
drive it came with.

So we have a Mac Studio now, which is a low-end UNIX workstation.
They aren't cheap, with a price on the order of magnitude of what
I paid for my Linux workstation.

Eyeballing the results, it looks like it renders Fooocus images about 1/4
the speed of my Linux workstation.  I thought it would be a lot
faster, but I guess pytorch doesn't support hw acceleration yet -- or
the hardware on the Mac is no match for an RTX 3090ti w/24G of memory.

You can say what you want, but I'd much rather use my Thelio
than the Mac Studio.  YMMV.

--
-v System76 Thelio Mega v1.1 x86_64 NVIDIA RTX 3090 Ti
   OS: Linux 6.11.10 Release: Mint 21.3 Mem: 258G
   "Mary had a little RAM -- only about a MEG or so."