Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
It would also require Apple to develop, manufacture and stock different "base" modules with system boards using consumer i5 / i7 CPUs and those using Xeon CPUs. You would then also have multiple BTO trees dependent on those architectures and what they support in terms of RAM type and capacity, PCI lanes and external connectivity (both with other Apple modules and third-party devices).

I do not see Apple wanting to support that level of complexity internally nor expose it to their customer base externally.
 
  • Like
Reactions: fuchsdh and Aldaris
Apart from TB not being up to the task , the costs for a solution with seperate modules would be ridiculous .
Love the idea, but the prices even for TB externals havn't come down to a reasonable level after years .
Apple has not shown many signs of properly supporting , much less developing their own external TB hardware in as many years either .

YEH ... I always thought they started with a modular version of the Mini, and ended up with the opposite, the Trash Can.

Shame about Thunderbolt not getting support. Especially since its normally supplied with today's notebooks, along with its USB 3.1.

Another type of Mac that has appealed to me for years ... is a Mecano Mac. Hey my idea, but I bet its been thought of.

Just have a set of Mecano type tools and build it yourself.

The "art" for Apple would be safety I guess ... and protecting their OS.

For a company that mostly builds closed items though, making something like that would be too much like thinking different.
 
For a company that mostly builds closed items though, making something like that would be too much like thinking different.

On the flip side, thinking different for the sake of thinking different gave us the 20th Anniversary Macintosh, the G4 Cube and the Late 2013 Mac Pro.

None of which proved to be practical.

Apple won't give us a PC OEM tower because they are Apple, but I sincerely hope what they do give us does not functionally stray to far from such a thing, even if it visually might.
 
This "box that resembles a PC tower" (no photo provided) looks interesting:
https://www.macrumors.com/2018/04/27/apple-ar-vr-headset-8k-displays/
"it would connect to a "dedicated box" using a high-speed short-range wireless technology called 60GHz WiGig. The box would be powered by a custom 5-nanometer Apple processor that's "more powerful than anything currently available" and similar to the custom chips that Apple will use in future Macs. At the current time, the box resembles a PC tower, but it "won't be an actual Mac computer."
 
Last edited:
It would also require Apple to develop, manufacture and stock different "base" modules with system boards using consumer i5 / i7 CPUs and those using Xeon CPUs. You would then also have multiple BTO trees dependent on those architectures and what they support in terms of RAM type and capacity, PCI lanes and external connectivity (both with other Apple modules and third-party devices).

I do not see Apple wanting to support that level of complexity internally nor expose it to their customer base externally.

Everything involves trade-offs, and I suspect that the one that Apple finally re-awoke to is that "Hardware Specs Matter", with the perspective that working a bit more upfront on a flexible base design has payoffs down the lifecycle when it comes to the later generation derivatives of that base design, as opposed to a "One Off" design that has to be figuratively 'entirely' thrown away in order to conduct any update.

For example, even the 6,1 Mac Pro had different CPUs that could be plugged into its motherboard's CPU socket, and one of the (few) things they did to try to refresh the product line was to replace a 4-core CPU with the 6-core. In essence, this is because the design's motherboard was "modular" because of the CPU socket.

Granted, there's more to the design details than merely the CPU socket .. there's thermal, issues of bottlenecking, etc, but in the end, they sold the 6,1 Mac Pro with a variety of CPUs and then were able to 'refresh' the tcMP (however mildly) a few years later to a (slightly) different CPU selection group.

RAM is a similar 'modular' beast when there's a slot for it instead of being soldered on. Ditto for HDDs/SSDs.

And sure, there's going to be OS & Software support required for changes such as a newer (or previously too expensive) CPUs, but these can be planned & scheduled ... and in the big "enterprise" level picture, its cheaper to add & test some new code than it is to design, build & test a new motherboard in addition to the same OS code revisions.

------

In any event, as more & more time slips buy without a shipping product, the more that I'm convinced that much of Apple's lethargy here is that they're trying to figure out how to be 'modular' from their own manufacturing flexibility perspective while keeping the changes cheap (for them) but while not maximizing their use of existing industry standards (such as PCIe), because doing so results in an erosion of profits because of losses to DIY'ers.

And sure, we can go point to alleged statistics that say the likes of "90%" (or whatever) of Mac users never revise their hardware, that's a pretty darn nebulous claim when you start to really look at it. For example, since over 80% of Mac customers today are using laptops which can't be changed, while the statistic is pedantically correct that "only 10% of all users" revise hardware, its ignoring that 80% of them can't make changes even if they wanted to. As such, the 10% is all contained within the 20% who can make changes ... and 10%/20% = half of them. Suddenly, the metric isn't so dismissive anymore, isn't it?
 
I've never owned a Mac Pro, but the older aluminum classic tower design was always the most attractive looking computer to me. If Apple released this same design in an up to date fashion with modern, respectable specifications I would honestly begin saving for it as soon as I heard the news and use it for the following decade.

I guess the lack of a release here is another reason users are so inclined to build a hackintosh (another being price of course). I want a traditional desktop, but I don't want it to be Windows, and while I could get a hackintosh to work, I don't want the feeling of walking on eggshells when trying to perform updates and the like. If a supported, stable, updated "old school" Mac Pro were released I think it would do very well.
 
Everything involves trade-offs, and I suspect that the one that Apple finally re-awoke to is that "Hardware Specs Matter", with the perspective that working a bit more upfront on a flexible base design has payoffs down the lifecycle when it comes to the later generation derivatives of that base design, as opposed to a "One Off" design that has to be figuratively 'entirely' thrown away in order to conduct any update.

For example, even the 6,1 Mac Pro had different CPUs that could be plugged into its motherboard's CPU socket, and one of the (few) things they did to try to refresh the product line was to replace a 4-core CPU with the 6-core. In essence, this is because the design's motherboard was "modular" because of the CPU socket.

Granted, there's more to the design details than merely the CPU socket .. there's thermal, issues of bottlenecking, etc, but in the end, they sold the 6,1 Mac Pro with a variety of CPUs and then were able to 'refresh' the tcMP (however mildly) a few years later to a (slightly) different CPU selection group.

RAM is a similar 'modular' beast when there's a slot for it instead of being soldered on. Ditto for HDDs/SSDs.

And sure, there's going to be OS & Software support required for changes such as a newer (or previously too expensive) CPUs, but these can be planned & scheduled ... and in the big "enterprise" level picture, its cheaper to add & test some new code than it is to design, build & test a new motherboard in addition to the same OS code revisions.

------

In any event, as more & more time slips buy without a shipping product, the more that I'm convinced that much of Apple's lethargy here is that they're trying to figure out how to be 'modular' from their own manufacturing flexibility perspective while keeping the changes cheap (for them) but while not maximizing their use of existing industry standards (such as PCIe), because doing so results in an erosion of profits because of losses to DIY'ers.

And sure, we can go point to alleged statistics that say the likes of "90%" (or whatever) of Mac users never revise their hardware, that's a pretty darn nebulous claim when you start to really look at it. For example, since over 80% of Mac customers today are using laptops which can't be changed, while the statistic is pedantically correct that "only 10% of all users" revise hardware, its ignoring that 80% of them can't make changes even if they wanted to. As such, the 10% is all contained within the 20% who can make changes ... and 10%/20% = half of them. Suddenly, the metric isn't so dismissive anymore, isn't it?
even with the Mac users never revise their hardware line. Having more choice at buying time is good.

The new mac pro did poorly with that and the IMAC pro start point is to high.

Some people need more cpu power but don't need an TOP $$$ range video card or really don't need 1 TB of high $peed Fla$h.
 
I do not see Apple offering Core family and Xeon family Mac Pro models, since the iMac 5K already covers the Core family.

If Apple chooses to offer only a single-CPU option, I would imagine it would be Xeon-W like on the iMac Pro. If they offer multi-socket options, then things could get very expensive since I would imagine they would use an LGA 3647 solution with Xeon SP since those can support single and multi socket on a single platform.
 
I do not see Apple offering Core family and Xeon family Mac Pro models, since the iMac 5K already covers the Core family.

If Apple chooses to offer only a single-CPU option, I would imagine it would be Xeon-W like on the iMac Pro. If they offer multi-socket options, then things could get very expensive since I would imagine they would use an LGA 3647 solution with Xeon SP since those can support single and multi socket on a single platform.
Motherboards are cheap to design, there's no reason Apple couldn't do what poor companies like HP and Dell do - build a chassis and give options from Core i7 to Xeon SP.
 
[T]here's no reason Apple couldn't do what poor companies like HP and Dell do - build a chassis and give options from Core i7 to Xeon SP.

Of course there is: greed.

Apple is too margin-focused to do anything that might, let alone would, reduce their margins.
 
  • Like
Reactions: OS6-OSX
Motherboards are cheap to design, there's no reason Apple couldn't do what poor companies like HP and Dell do - build a chassis and give options from Core i7 to Xeon SP.

Well they could scrap the iMac, the Mini and the Mac Pro and just offer one mid-tower case design family.

Maybe call it the Macintosh III.
 
I think, besides Cook Storage-Manager mind (force iMac Pro sales), a main factor conspiring to delay the Mac Pro is the GPU market volatility, Imagine Apple launches it with GPU options at 1500$ each and then the GPU maket crashes and Apple needs to sell those custom GPU cheaper or see MP users to switch to cheaper solutions.

Even the iMac pro/5k has no economical sense, since Apple actually almost subsidizes the 5K display (iMac 5K is even cheaper than a stans alone 5K display), Apple could earn much more by selling it modular.
 
Of course there is: greed.

Apple is too margin-focused to do anything that might, let alone would, reduce their margins.
Well they could scrap the iMac, the Mini and the Mac Pro and just offer one mid-tower case design family.

Maybe call it the Macintosh III.
1000 words:
mobo.jpg

If Asus+MSI+Gigabyte have almost 900 different Intel mobos - it's complete BS to claim that Apple's margins won't let them make two mobos.

Complete fanboi BS that ignores the facts. Don't be apologists.
 

Attachments

  • mobo.jpg
    mobo.jpg
    53 KB · Views: 153
Last edited:
  • Like
Reactions: Mago and Synchro3
even with the Mac users never revise their hardware line. Having more choice at buying time is good.

Sure, if that choice is offered to the customer. Apple can minimize how many CPUs it chooses to sell, even if it changes which CPUs these are every ~12 months. The modular approach of a CPU socket allows them to use the same production line with a simple parts swap (and minimal reprogramming of the pic-n-place robotic arms).

The new mac pro did poorly with that and the IMAC pro start point is to high.

Some people need more cpu power but don't need an TOP $$$ range video card or really don't need 1 TB of high $peed Fla$h.

Ironically, on the Quarterly call, Tim Cook just claimed that their product pipeline is the best its ever been.

In any event, what the cMP tower did that other designs have struggled with is to be tailorable to the diversity of 'pro' use case needs ... as you point out, not everyone needs several performance metrics all maximized, resulting in a steep $5K starting point before they even start to add to that to tailor the base product to their specific use case needs.


-hh
 
1000 words:
If Asus+MSI+Gigabyte have almost 900 different Intel mobos - it's complete BS to claim that Apple's margins won't let them make two mobos.

Complete fanboi BS that ignores the facts. Don't be apologists.

I think you missed the meaning.

It would of course be trivial for Apple to redesign a second (or third, or a hundred) motherboards. However, every single SKU in inventory reduces margin by raising costs. And Apple has a unblemished history under Cook of doing everything possible to raise and/or maintain margins. Inexcusable for the end user, but exactly correct for manufacturing math.


As a PS, we also have zero idea what margins Asus, MSI, and Gigabyte maintain.
 
However, every single SKU in inventory reduces margin by raising costs.
But, what if the extra one or two SKUs increase sales?

Badda Boom!!
[doublepost=1525217165][/doublepost]
As a PS, we also have zero idea what margins Asus, MSI, and Gigabyte maintain.
Apparently large enough to have been in business for a long time.

Stop apologizing. Apple could sell each Mac Pro for 50% of its cost without materially affecting the profit margin.
 
But, what if the extra one or two SKUs increase sales?

Badda Boom!!
[doublepost=1525217165][/doublepost]
Apparently large enough to have been in business for a long time.

Stop apologizing.

Who is apologizing? Clearly you're confusing me with someone else.

And while you appear to have plenty of IT experience, that does not translate to the logistics of manufacturing.

Good day, ma'am.
 
And what are your "bona fides" in manufacturing logistics? ;)

None. ;)

My spouse, however, is the vp of finance for an electronics manufacturing company, and teaches in the local university's mba program. The more SKUs involved, the harder it is to maintain a given margin.

Why you're interested in attempting to debate something well-settled is, frankly, beneath you.
 
My spouse, however, is the vp of finance for an electronics manufacturing company, and teaches in the local university's mba program.
WOW - pure BS. ;)

So, Asus has over 1000 different mobos on Newegg - but it would kill Apple's margins to make a second one.

It boggles the mind that you can believe that....
 
WOW - pure BS. ;)

So, Asus has over 1000 different mobos on Newegg - but it would kill Apple's margins to make a second one.

It boggles the mind that you can believe that....

Sigh.

Are you having a bad day?

As I said earlier, good day, ma'am.
 
You should have stopped here, and not tried to embellish what your husband has done.
Not trying to get too involved here... but SKU count hasn’t effected the iPhone lineup, or the watch lineup. So what’s the harm in an i5, i7, i9, Xeon desktop Mac?
 
Thought I’d check in on this thread, and I can’t believe all the good stuff I’ve been missing!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.