Wow, your reading comprehension is lousy...!
I clearly stated how the new 2019 modular Mac Pro would be the LAST great Mac BEFORE the transition to Apple-designed CPUs...
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MP 7,1 Waiting for Mac Pro 7,1 A1991 (no more)
- Thread starter Mago
- Start date
-
- Tags
- macpro 2016
- Sort by reaction score
Wow, your reading comprehension is lousy...!
I clearly stated how the new 2019 modular Mac Pro would be the LAST great Mac BEFORE the transition to Apple-designed CPUs...
I guess you are right that I didn't comprehend the point of your comment.
I seriously don't understand what "The Transition" has to do with waiting for the 2019 Mac Pro unless you think it will be the LAST Mac Pro not based on an Apple CPU, and don't want to miss out on it because you bought something else before then instead.
Would you not wait and instead buy something else now if you thought "The Transition" wasn't coming in 2020? Why would that be?
What I mean is that if I were in the market for a new Mac (which I am), I would pass on the new SG mini in favor of the forthcoming modular Mac Pro...
The new modular Mac Pro may very well be the last Mac to have Intel CPUs (okay, Apple may update the iMacs before then, but the non-upgradable GPU & monitor are a long-term deal breaker), which would give me a solid machine to 'ride out' the transition period from Intel to Apple-designed silicon, ensuring legacy software compatibility while everything makes the leap from Intel to ARM...
After a few years, once the ARM Macs are 'all good', then I could feel confident in making the switch from "the last Intel Mac" to the 'new hotness' (to quote Agent J from MiB II)...
And I never said "The Transition" wasn't coming in 2020...
I am quite excited by "The Transition" and look forward to what Apple might do with systems entirely of their own design...
If there is one thing Apple avoids like the plague, it's hardware and firmware support , unless it's absolutely unavoidable . Or any support really , it's a money dump .
So why would they open that can of worms by owning a CPU supplier ?
So I guess my reading comprehension wasn't bad after all.
As I said to you the first time, that isn't going to happen, unless it's the LAST Mac Pro ever.
As I said to you the first time, that isn't going to happen, unless it's the LAST Mac Pro ever.
Why? Are all the Pro users going to leave macOS & Apple hardware because of a CPU change...?
You make reference to issues with the transition from Power PC to Intel; that was rough, but it did not stop pros from making the switch to Intel Inside Mac Pros, why would a change to ARM be any different...?
I seriously doubt Apple has plans to transition their entire lineup over to ARM, but will decide to kill the Mac Pro...
Not crazy? That would indicative that the folks in charge of Mac product management and the Mac Pro in particular were Kings of the Bozos. That amount of exceedingly poor rationalized , non-linear nonsense at the top of the Mini BTO selections is off the charts. That zero rational reason to use that to form the baseline of entry Mac Pro pricing. None. The only "rational" reason for using that as a baseline for Mac Pro pricing is if Apple wanted the Mac Pro to fail. (e.g., "Let's price it stupid so hardly no one buys it so we can kill it because 'no one' is buying it. ).
The Mini 2TB SSD is priced so high you can buy a external M.2 SSD enclosure and a M.2 2TB and still have hundreds of dollars left over. Putting that as a 'floor' under Mac Pro prices is ludicrous. There is a small subset of folks buying Mini's who are using "other people's money" to buy that BTO configuration because the revenues in the business context they are in can pay for it and have a corner case where need all the storage inside of one box. But that can't the same rational used for the baseline Mac Pro customer.
The iMac Pro is priced clear of much of the iMac BTO zone in part because it is largely an iMac. (even there it is highly suspect move; given the somewhat chronic discount pricing on the iMac Pro entry config that's a minor problem). That is a corner case exception. ( I suspect Apple was also nervous whether the iMac Pro would work and was looking to gamble a bit pricing to hit breakeven sooner even if that damaged long term viability. They could tweak version 2.0's baseline pricing if got confirmation was viable. if wasn't at least covered costs of the experiment). However, the rest of the Mac product line up has overlapping BTO zones. Mini overlaps iMac , MBA overlaps MacBook Pro (and MacBook). Heck even MBP 13" overlaps MBP 15". Apple does not have a general "BTO ranges must not overlap" rule. that's largely because that would be a bozo rule. So they don't.
Furthermore, The baseline entry Mac Pro is extremely likely targeted at folks who don't want an iMac and probably don't want a Mini either. If they weren't buying an iMac 'period' then pushing them out of the iMac BTO zone does a whole lot of nothing , but drive a boatload of potential customers off. Large groups of folks who spent 3-10 years not buying iMacs (and/or iMac Pros ) are extremely unlikely to change their minds. Is Apple going after the Xmac crowd? No. They aren't going to strip down the Mac Pro so it drops into the sub $2K zone.
The base price will probably be relatively close to where it is now for many of the same reasons the old systems were in that zone. That is about the range of budget that a significant number of targeted users have. The Mac pro isn't a different config that the iMac form factor so it could start at 6 cores even though the iMacs will probably be at 6 cores at that point also. There is a set of folks who also can deal with a lower mid-range GPU ( could put a 'Polaris' equivalent class in there dump tons of BOM costs into the default GPU).
Plenty of RAM in a Mini with just two so-DIMM slots? If the Mac Pro has 4-8 ECC capable DIMM slots the Mini doesn't compete on "plenty of RAM". The mini has more capacity than the old mini. The capacity on the approaching 10 year old Mac Pro is 128GB. That is fully 100% larger than the 2018 Mini's top end of 64GB (and non ECC).
Yes, there will be more folks who 'opt out' of a Mac Pro at the bottom entry level end. But that doesn't mean have to make the entry model more expensive to "push" folks that way. Some of those folks are going to want two ethernet port ( one for internet , one for storage or admin network . ) More than a few are going to want a far more viable GPU. The Mini is decent for pedestrian GPU workload, but largest sized screens and/or heavier workloads it isn't a good match at all. ( and 'get an iMac' isn't going to work for the folks who viscerally don't like iMacs. ). it is a poor match to the 2013 Mac Pro which was more ideally targeted to folks who had two GPUs worth of workload. Even in single GPU performance the relatively ancient MP 2013 GPUs walk all over the 2018 Mini GPU. For folks who put an equivalent or higher weighting on the GPU versus the CPU, the Mini is somewhat of a 'dog' choice. It isn't good. (that is why many of its higher BTO configurations are going to have a minimized fratricide impact on the iMac; let alone the Mac Pro core space. )
Folks who bought the Mac Pro because it was a headless, literal desktop and but somewhat overpaid for mainstream desktop performance needs, the new Mini is a probably a better match. I don't think that was the primary core of the folks who bought the MP 2013 though.
I can tell you are very heated on this subject and that was a whole lot of words with very little actually said. I’ll just note that your concluding paragraph here agrees with my post which you appear to have found so objectionable.
I never said anything about the “primary core” of 2013 Mac Pro buyers, but rather observed that the people who were buying the base configuration would be adequately served by today’s high spec BTO Mini. This is a conclusion you appear to agree with: “the new mini is probably a better match.”
It stands to reason that the base configuration of the 2019 Mac Pro will not need to be neutered to that level if Apple has a solution for them already. I take this as a hopeful sign that the new Mac Pro will instead cover a higher range than the outgoing 2013 and possibly/likely even that of the old cheese grater builds. This would be more in line with the full range or offerings from HP and Dell, if you consider those workstation products to be the competitive landscape that Apple is in.
That's a bit too narrow. They are targeted to be computational focused cards. There is only one minor output connector on the 'outer' edge of the card. ( not targeted at display output ). However, in the primary display GPU role Apple wouldn't be shipping the card. ( open question whether AMD tossed all the several display output stream handling for compute in these variants. Or later option for mainstream is their next gen gets stuck. )
As a secondary, computational GPU it is targeted at workstations. Some folks doing code development will have one in a workstation to run incremental code/debug/test iterations before deployment to datacenters and cloud services. that's the kinds of stuff that Apple will push themselves out of if don't have a empty slot in the new Mac Pro. There aren't going to be more developer deployments than "service providing" deployments, but won't be 'close to zero' either. Most will be lLInux workstations, but is a space Apple could participate in if put in some minimal effort. ( MI60 and MI50 only officially supported on Linux x86_64 ... not even Windows. )
The primary, boot display card? Yes, not even close to being appropriate. New boot card should be a GPU that is nominally oriented toward driving 6 displayPort streams. ( not backsliding from the MP 2013 although all of those don't have feed into Thunderbolt at least 4 would be minimal. Same baseline as the iMac Pro).
An optional BTO secondary compute card... that is a different story. The fallacy that Apple shouldn't be holding onto to is the fallacy that the boot display GPU and any secondary compute GPU have to be 'identical twins' . That's goofy (which was one of the 'corners' the MP 2013 backed into) .
However, as an add-in option it shouldn't hold up the launch in any significant way if it was the lone hold out on readiness.
Uhhh...
That Core i9 from MBP Is the Core i7 from Mac Mini, but Mac Mini will maintain higher clock speeds, because on non restricted power delivery...
What for? AMD has Semi-Custom division that can easily make special custom chips for Apple, specially with chiplet architecture.
Building an APU from GPU chiplets, and CPU chiplets for BGA platform would be easy for AMD, and profitable for them.
That would make more sense...
But if Apple is transitioning to ARM APUs of their own design, that could be interesting (once the software hiccups are quelled)...!
Interesting, if you want an Iphone with a keyboard and a big screen.
ARM-based systems would be the end of the desktop OS as we know it, and the end of many of the third party apps that help keep OSX relevant.
I'm not sure I understand why people assume Apple would stick with ARM if they were to design their own Mac CPU's?
Frankly, I'd like to think they'd start with a clean slate and really build a modern non-mobile processor (PPC MkII, lol). Heck, they could even build their own x86 CPU, although I really hope they don't.
The thinking is that if the A12 can achieve that kind of performance on what is it, 5W?* Imagine what they could do without TDP restrictions.
*edit:
I'm having difficulty finding a definitive answer to the A12's TDP question. If I find it, I'll amend this post.
edit 2:
This article from Ars Technica has Schiller and Anand explain some of the rationale behind their processors: https://arstechnica.com/gadgets/201...e-ipad-pros-a12x-system-on-a-chip/?comments=1
I found this quote from Anand particularly interesting:
"Fundamentally, the reason we built the chip, is in service to the product's vision and its ambition," Shimpi said. "At the end of the day we want to make sure that whatever vision we have set out for the thing, if it requires custom silicon, that we're there to deliver. For a given form factor, for a given industrial design in this thermal envelope, no one should be able to build a better, more performant chip."
From this I'd conclude they'd take a look at what the Mac does and design the chip specifically for that use case.
Or that's what I hope, anyway!
Frankly, I'd like to think they'd start with a clean slate and really build a modern non-mobile processor (PPC MkII, lol). Heck, they could even build their own x86 CPU, although I really hope they don't.
The thinking is that if the A12 can achieve that kind of performance on what is it, 5W?* Imagine what they could do without TDP restrictions.
*edit:
I'm having difficulty finding a definitive answer to the A12's TDP question. If I find it, I'll amend this post.
edit 2:
This article from Ars Technica has Schiller and Anand explain some of the rationale behind their processors: https://arstechnica.com/gadgets/201...e-ipad-pros-a12x-system-on-a-chip/?comments=1
I found this quote from Anand particularly interesting:
"Fundamentally, the reason we built the chip, is in service to the product's vision and its ambition," Shimpi said. "At the end of the day we want to make sure that whatever vision we have set out for the thing, if it requires custom silicon, that we're there to deliver. For a given form factor, for a given industrial design in this thermal envelope, no one should be able to build a better, more performant chip."
From this I'd conclude they'd take a look at what the Mac does and design the chip specifically for that use case.
Or that's what I hope, anyway!
Last edited:
Two big reasons why Apple won't go to a custom CPU in the Mac any time soon... Adobe and Microsoft. Neither one has any interest in supporting another architecture, and the Mac needs both. After Apple moved to Intel, Microsoft Office moved a lot closer to feature-parity with its Windows version, and companies started allowing employees first to bring Macs from home, then to request them. I believe IBM (for its own employees) is now the single largest Mac customer!
But they are both already supporting ARM architecture; Microsoft with Office 365 & Adobe with a slew of products for iOS, including Lightroom, Photoshop (forthcoming), Premiere Rush, etc. ...
This. It's blatant that Adobe have developed Photoshop for forthcoming ARM Macs, there is no way that they'd sink all that dev time into Photoshop just for the iPad. I've been a graphic designer for 25+ years and no one I've spoken to in my industry is interested in using Photoshop on a tablet, beyond very minor edits. This is all about Adobe porting Creative Suite to ARM based Macs and having it ready for the first generation machines. It is coming.
Regarding ARM-based Mac Pros...
Think about the A12X Bionic APU in the Gen3 iPad Pros...
No active cooling, just passive cooling with dissipation to the iPad chassis itself...
Now, think about what could be done with active cooling...
Now think about two or four of these APUs working in one compact chassis...
Think about the A12X Bionic APU in the Gen3 iPad Pros...
No active cooling, just passive cooling with dissipation to the iPad chassis itself...
Now, think about what could be done with active cooling...
Now think about two or four of these APUs working in one compact chassis...
C
RISC could easily replace CISC again. We are a long way from 2006.
x86 is a dinosaur with a **** load of legacy crap no one has any use for. It is even holding technology back at this point in time.
RISC could easily replace CISC again. We are a long way from 2006.
Going from a PowerMac G5 to an Intel MP was a significant step forward in performance back then , amongst other advantages . Not to mention the move had been requested by virtually everyone in the user community .
The way I remember it, the transition was fairly smooth too, at least for me .
Easier than moving from Snow Leo to Mavericks and beyond, that's for sure ...
Backwards compatibility was still a thing in those days , and ever important for pros, but something Apple has abandoned since then .
Apple also used to have a pretty solid foothold in the creative business and was still going places in that industry , which isn't the case anymore .
Today, Apple is not in a position to switch CPU providers with all the issues that might come with it, with no huge gain for the customers - and there would be no such thing .
Did I mention recent OSX backwards compatibility ?
For the past few years, there's been one request only, which is for Apple to stop being silly and to offer an updated cheesegrater .
How would Apple manage the transition?
I mean, when they did 68k>PPC, they had an emulator in the Nano Kernel and Fat binaries. When the did PPC>x84 the had Rosetta.
Thing is ARM can't emulate x86 very well at all, I just don't think an ARM CPU would be good enough to run Pro level software durning the transition.
Leaving the only real option of an ARM coprocessor, with an x86 CPU. Sad thing is, people would figure out real quick that the same code run on ARM, would be a lot slower than x86.
The ARM cpu is good at what it is good for, performance per watt in low powered devices. Desktop Pro machines don't need to fit in those boxes, they are all about maximum performance per dollar.
I understand Apple's longing to move everything to one codebase, and wanting to control their devices end to end, however I just don't think they will be able to convince Pro users to make this switch, but time will tell.
Care to elaborate ?
From where I am standing, I can only see prices going up and the products barely getting better for what my money buys .
Compare cars from 2006 to current ones, compare their price, and you might understand how progress is supposed to work .
I heard that these compiler things can take your source code and spit out binaries for different systems. If only Adobe, Microsoft, et al. could do stuff like that, maybe they could port programs to run on different hardware. I guess they would have to have compiled code--not assembler-- and with foresight structure it for porting to different systems, which would probably be beyond the state of current technology. Oh well, we can dream of the future.
Microsoft already offers a Windows 10 ARM edition (both 32-bit & 64-bit versions), so there wouldn't be a major problem adding Bootcamp Windows 10 to an ARM-based Mac. Windows 10 ARM edition includes an x86 emulator, but AFAIK can only run 32-bit Windows programs, not 64-bit programs, presently.
Didn't Apple move away from "custom" CPUs and to Intel as they were unable to compete? Apple moving to ARM sounds like a move back from where they came. Designing and producing competitive CPUs is hard. We've seen the industry consolidate into a small number.