It's competing against machines such as the Boxx dual-CPU workstations, which start at $8000 [1]. I can see movie studios and such shelling out the big bucks to buy these for their 3D modellers, but only the super rich would buy one for themselves.
[1]: https://github.com/KhronosGroup/MoltenVK/blob/master/Docs/Mo...
I don't know much more about Metal beyond that though.
What do 3D modelers use CUDA for? I know it gets used for machine learning, but I'm not familiar with what else.
So, my bold prediction: Apple is going to shred Nvidia to pieces within the next 10 years.
CUDA/OpenCL is all about writing super parallelism code to be run on the GPU instead of CPU, so many trivially parallel problems are usually outsourced to it.
Source: https://everymac.com/systems/apple/mac_pro/mac-pro-cylinder-...
In terms of just GPGPU performance Vega is pretty competitive against Nvidia. It's not nearly as one-sided as the gaming space is. That is, in fact, the one saving grace of the Radeon VII in the first place vs. the RTX 2080 - its stronger professional workload capabilities.
That $8000 base model Boxx workstation has 20 cores instead of 8, 50% more RAM, and twice the storage.
You would get something like this instead of consumer hardware because man hours cost real money. At a previous job they wanted a VR PC, so someone speced out a bunch of high end parts and a case with flashy lights. A couple days of my salary went to putting it all together. That was the first time I had to RMA a motherboard. When I asked they didn't think to confirm the GPU would fit in the case and luckily it did. It also took a couple weeks longer than it should to get up and running because of things like this.
Some specialty software is only certified for specific hardware. You can probably deviate, but it's like building a hackintosh. You might not be able to upgrade it and the man hours will likely kill any benefit.
I also see a lot of managers forget that artists aren't always tech savvy. Honestly, you don't pay them to be tech support, anyway. You want to pay a known quantity and get improved return. It can be a huge time suck to nickel-and-dime when you're tying to manage large projects.
However, using consumer grade hardware has its place.
Honestly, Apple wasn't what I had in mind when I was replying. Personally, I've had to deal with HP z-series workstations compared to various cobbled together machines. Having machines where the series will be in production for a decade, data sheets are readily available, fewer tools needed to open it and install stuff saves so much time and allows less downtime and faster turnaround.