Does a Hacktinosh Point to the Solution to Apple’s Mac Pro Dilemma?
May 2nd, 2017I don’t have to repeat the obvious. Better late than never, Apple finally came to realize that its 2013 revision to the Mac Pro was a big misfire. Well, they didn’t use those terms, but it was the polar opposite of the previous version. Where the so-called cheese grater Mac Pro made it simple to add extra stuff inside, the Mac Pro pushed most expansion outside. There were lots of ports, but no place inside for extra drives or expansion cards.
I’m not making a point of the lack of twin CPU ports, because Intel has solved that problem by adding more cores to its Xeon silicon. So if you need 12 cores, it’s better to have them all on one chip than two. You get the picture.
Apple isn’t going to say how long it took to get a clue about what it did wrong. You’d think complaints from professional users would be enough, or maybe Apple thought they’d eventually get used to it. But in saying they are working on an all-new model, Apple executives did claim to have talked to pros. Sometime. Somewhere.
Why’d it take so long to do the obvious? I wouldn’t care to guess why Apple had such a bad case of tunnel vision. The point is that the message was finally received loud and clear, Apple has more or less made its apologies for being so obtuse, and has promised that a hardware team is working on the new model.
Now some might suggest that it would be extremely easy to deliver that replacement, and professional Mac users shouldn’t be forced to wait at least a year for satisfaction. Maybe Apple should look to the Hackintosh community for inspiration.
Or at least that’s what is implied in a fairly detailed article on the subject in Ars Technica. We have perhaps thousands of people taking off-the-shelf PCs and installing macOS on them by jumping a few hoops. And it does appear the hoops have become fewer as the hobbyists get a better handle on the process.
So far Apple has treated this practice with benign neglect. It’s not as if these people are going out and selling undocumented macOS clones. If they did, Apple would stop it — and they’ve done that before. What the Hacintosh demonstrates is that people are so dedicated to the platform that they are willing to spend time and money assembling a Mac that meets their needs.
But these are not unique designs. Desktops are mostly regular PC minitowers, perhaps in the tradition of the ancient Power Computing boxes. It’s of secondary importance, since the real creativity is in choosing the proper internal components that are known to be adaptable for use in a Hackintosh. But it also means these boxes are easily expanded with extra drives and expansion cards, meaning they are built in the spirit of the original Mac Pro.
So the simple solution would be for Apple to restore the cheese grater model with new parts. This would be a simple solution that shouldn’t cost a whole lot in development dollars. It could still have USB-C/Thunderbolt 3 ports, and even support for a keyboard with a Touch Bar.
Sure, Apple could perhaps refine the cooling system, and even make it slimmer and lighter without detracting from its fundamental expandability. It could even come in dark gray or black if Apple can find the right color scheme and casing.
For now, Apple is expecting customers to buy the current Mac Pro at a lower price to make do, or wait. Some might just build their own Hackintosh instead.
Clearly, Apple isn’t taking a year to refine a Mac Pro minitower unless there are other design factors involved. But that takes us into a wide area of speculation into what a proper computer workstation should be in 2018 and beyond. Obviously Apple knows about the roadmaps for its silicon vendors. They know what Xeons will come from Intel — and it’s not as if this information is necessarily top secret. They know what AMD is doing to compete, if anything, plus the future design goals for its graphics chips. They know what’s coming from NVIDIA.
But what if Apple plans to build a Mac Pro with its own graphics hardware? Obviously that is being done for future iPhones and iPads, witness the recent news that they are phasing out use of GPU intellectual property from Imagination Technologies. Apple has already hired its own GPU engineer, but why should it be restricted to iOS? What about macOS?
It’s not the same deal as switching from Intel to ARM for Macs. Graphics technologies span platforms, so if Apple has a better idea, more tightly integrated into the macOS platform and its own needs, maybe going its own way is the better idea. So why not start with a state-of-the-art computer workstation that can go toe-to-toe with graphics from Intel, AMD and NVIDIA?
Maybe the delay is not so much the form factor as completing and testing its own GPU design. I suppose if it all doesn’t come together in time, meaning by next year, Apple could always go to its current chip partners and continue working on its home-built solution for use at a later date.
After all, why should it take so long to design a modular Mac Pro?
Print This Post
If Apple did design a new Mac Pro that only worked with the hypothetical Apple proprietary GPU then this would likely cause a lot of anger and grief.
Even now after years of Apple only using AMD based discreet GPU designs along with some Intel integrated GPU designs many professional Mac users are using apps that are still optimised for Nvidia GPUs which Apple have not shipped for years. (Cough – Adobe – cough.)
Can you really see Adobe re-writing all their software to work with a proprietary Apple GPU? They still have very limited support for Apple’s Metal API and still also poor performance on AMD GPUs in a Mac.
In any case Apple’s recent press event over the future of the Mac Pro clearly said it was going to be an expandable machine and this would seem to make it pretty certain it will support standard PCIe cards which is what professional users have been screaming from the roof tops for.
Where a hypothetical Apple GPU might have a justification might be as a replacement for Intel Integrated GPUs. As an example to add hardware support for the H.265 codec and thereby make Macs as good at handling 4K video as they are for 1920×1080 high definition in H.264.
Obviously one of the considerations Apple would confront, though some level of basic one-click support in Xcode might help. Or use the custom GPU as some sort of coprocessor.
Peace,
Gene
Hackintosh will always be an alternative, but licensing MacOS to a controlled/select group of manufacturers committed to high-end, frequently refreshed hardware should to be considered. I own several iMacs, one a factory loaded 5K I find exceptional, but not upgradable with more advanced video chips/cards available.
Apple’s “updated” 2013 MacPro was based on style, the event commenting on its failure 4 years later was reactionary, instead of an “updated” 2017 introduction. Clearly, MacPro’s are no longer a priority, nor will they be in Apple’s future. Licensing MacOS should be considered, before we find Windows 10, “just works” in all that superior 2017 hardware out there, today. Style, thinness and smaller need to be considered after functionality.
i think it’s interesting the features that hackintosh builders (mostly) seem to value are single thread speed over number of cores (probably because of adobe) and expandability. there doesn’t seem to be huge appetite for the premium priced xeon chips apple use to differentiate the pro, outside of servers and rendering video (which i guess don’t really need the niceties of macOS anyway). The Imac pro is rumored to use a xeon variant which is just a i7 with ECC ram, which really sounds like something that’ll be called out endlessly by critics of apples pro credentials.