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
At the entry points of the laptops that has some relevancy. At the mid-top end of the Mac line up it is almost entirely irrelevant. For the Mac Pro 2013, those 'corners' were almost 99.99% self inflicted corners when it came to the CPU. Even with the corners they had trapped themselves in the could have moved down the Xeon E5 line to v3 or v4. For the GPUs, it was about 50+% self inflicted by sticking to the baselines of that design. Smaller than a Mac Mini desktop footprint was a problem not competitive system needed to actually "corner" themselves with, so particularly a component problem.
Neither in particular is the iMac Pro on CPU ( the regressed on RAM door but that is neither a CPU or GPU component fault in the slightest). The iMac ... even more self inflicted since the iMac Pro works with some sensible updates. The iMac's CPU fit in an even smaller thermal window. So does the GPU's. [ perhaps if Apple's quest was to make the iMac as thin as a TV to hang on the wall the form would play a larger role, that isn't particularly necessary at all. Hasn't been for last 10 years. Shouldn't be for the next 10 either. ]
The Mini ... I could imagine Apple would grumble a bit there but the is more fab issues with Intel than anything else. If had gone AMD it wouldn't have as big of an issue and would be even less by the end of the year. Again self infliction of OCD stuff than anything they are going to resolve with technology. Only here is about the point where the form factor should start to make some material difference.
For the laptops striving for the thinnest , lowest power dimensions practical then the Intel/AMD CPU-GPU combos are relevant.
Apple got "control" over being a Mac by putting the T-series in. That's getting back a bit to the 68K and PPC era when Apple had ROMs which were essential to Mac-ness. The fact those ROMs spanned 68K and PPC is entirely indicative that "control" really doesn't necessarily have much of anything to do with the instruction set, fab tech , or any of the "hocus pocus" folks trot out as an Apple, "control freak" driven objective.
If there was a broader and more powerful ecosystem to jump into ( like on the 68K->PPC and PPC->x86) then perhaps, but Apple jumping into a smaller, more brackish pond doesn't even help Apple really.
Lots of this is folks arm flapping about how the could blow large wads of money out of the Scrooge McDuck Apple money; not actually make money for Apple (and stockholders.).
Even intel has said “Apple is looking to drop intel”. I would imagine there has been some signaling between Apple and intel for intel to think that. We can keep on thinking oh xxx-lake is coming out next quarter but nothing has happened. Apples ‘pro’ software hasn’t seen any major attention, and neither has there hardware... so what else could it be? I’m not saying these are the only options I’m asking what other alternatives are there? Because both don’t seem to have a great pathway for pro/sumer/fessionals
I don’t know what’s going on in apples R&d anymore than anyone else in this forum, but reading what Apple has said and what Apple has been shipping I really tend to think Apple will either hit us with a massive “the Mac has been living a double life and we have a silicon that blows intel xeons-after everything we have learned with our powerful iPad Pros over the last 3 years we have developed a desktop silicon that beats everything on the market and we will be releasing it later this year”
Apple isn’t competing with anyone other than Apple-apple dropped the old iPhone cpu’s Because they saw that ‘they’ could do it better, why not here? Maybe it’s not ARM, maybe it’s their own x86 set who knows but Apple.
The argument that Apple is waiting for the next GPU or CPU at this point is laughable. You’ll jump over current offerings and still offer what was considered, by some, dated when it was released in 2013.
You mean "run x64 code" in emulation, right? x86 isn't that important anymore - it doesn't really matter how fast 16/32 bit code is.
Frankly if the performance penalty is on par with Rosetta the vast majority of people won't care at all. That's not necessarily the Mac Pro segment, but the next Mac Pro isn't going to be ARM so it doesn't really matter at this point.
Good points as always, sorry to cherry pick some quotes for readability . 
True .
I just never found it as difficult to adapt my workflow and gear to OSX updates as it is now .
That might well be just based on my personal experience and individual needs, of course .
Software and hardware vendors need an incentive, though .
Apple going more iOS and gadget centric, while neglecting the Mac line or taking pot shots with iMPs and touchbars and such nonsense doesn't create much motivation and perspective for 3rd parties .
That I can't quite follow .
Are you talking about Mac/GPU related issues only ?
Either way, most ambitious software made for or made available for OSX requires a reasonably powerful workstation Mac model to shine, doesn't it ?
Even if an MP is not strictly necessary for all use(r)s, surely a software house wouldn't want to waste resources on a Mac lineup, that significantly restricts the performance of demanding programs in its top models .
And what serious content creator wants to invest in a lineup that tops out at iMac level, or an ipad with a keyboard duct taped to it ?
It's not the number of MP sales, it's the general competence of the Mac line as a whole .
True .
I just never found it as difficult to adapt my workflow and gear to OSX updates as it is now .
That might well be just based on my personal experience and individual needs, of course .
Software and hardware vendors need an incentive, though .
Apple going more iOS and gadget centric, while neglecting the Mac line or taking pot shots with iMPs and touchbars and such nonsense doesn't create much motivation and perspective for 3rd parties .
That I can't quite follow .
Are you talking about Mac/GPU related issues only ?
Either way, most ambitious software made for or made available for OSX requires a reasonably powerful workstation Mac model to shine, doesn't it ?
Even if an MP is not strictly necessary for all use(r)s, surely a software house wouldn't want to waste resources on a Mac lineup, that significantly restricts the performance of demanding programs in its top models .
And what serious content creator wants to invest in a lineup that tops out at iMac level, or an ipad with a keyboard duct taped to it ?
It's not the number of MP sales, it's the general competence of the Mac line as a whole .
I think you're missing the point. The fact that they're only that far behind in a 5.9mm device vs. a large 27" desktop is quite the achievement. You don't think they could scale the cores and cache up to that TPU? Personally I think Apple is going to crush them.
[doublepost=1552401455][/doublepost]
True blue, you're in denial. Give it a couple more years.
You're probably right regarding socket 3647. Dreams of DP Mac Pros are likely just that. To an extent that's fine. We get so much on a single processor now that its probably not really worth it for Apple in particular.
Intel isn't the only x86 supplier. And they aren't the x86 supplier that is smoothly executing progress right now either. AMD makes semi-custom x86 chips for Sony and Microsoft. If Apple put money on the table to get one they could get some.
There is also little indication in these "drop Intel" rumblings that it was 100% complete dropping. Apple is throwing out all modems , CPUs, Thunderbolt controller , ... everything with an Intel badge on it out the window in 2020.
The modem part I believe far more than 100% drop from the Mac line up. However, that would be couple years away (at best). Intel's fab screw up has had more impact there than in the CPUs. If somehow Qualcomm and Apple bury the hatchet and Apple works out a deal for the price points they want ... Intel's modems in iPhones is toast. The move off of Qualcomm was not to something better/faster/lower power. Intel primarily has the business right now because they were not Qualcomm and were willing to throw gobs of money at tweaking to Apple specs.
Apple just moving the MacBook to a "hand me down" A12X/A13X would be a substantive drop in orders from Apple ( for a not so cheap CPU package). Also extremely likely a partial legitimization of the Quallcomm 8cx Windows systems coming down the road starting late this year on the Windows laptops size. That could also trigger even more substantive drop in orders. Let's say Macbooks are 600K per year in units. The Core m3 7Y32 is about $280. That's $168M in lost says for Intel if Apple just flips that one system. If 4 other big vendors also flip about the yearly run rate over to ARM that's $840M.
If Apple took the MBA off and that was 2M units/year that is another $580M. Add those up and at $1.4B. That's isn't going to put Intel into the poor house. But it is something to say is a significant possible issue. Apple just dropping Intel from just two "bottom end" laptops could be around $700M hit for Intel.
That would certainly send a "get your crap together" message to them to stop screwing up.
There is more than signaling going on. Intel is basically is at least 2-3 years behind on their roadmap. If the next Mac Pro ships with Cascade Lake 2066 socket chips in mid 2019 then probably a decent portion of that slide into 2019 was Intel's FUBAR. When the new Intel CEO took part of his initial messaging was that they needed to get much better at execution. It is about 99.9% likely that was a common message bubbling up from their largest customers.
It is somewhat likely Apple has stopped talking to them about a version of the "m3 / Y" class processor for the next MacBook. Constrained down to 8GB (maybe 16GB ) max RAM, no Thunderbolt, One port. 12 inch screen (like iPad Pro). Intel doesn't have something that is "better" if chasing the holy grail of thinnest.
Here is Intel's motherboard stab at "Always On, Always Connected" the worked up with HP.
https://www.anandtech.com/show/13434/intel-custom-amber-lake-y-with-lte-modem
and here is the iPad Pro board.
https://www.ifixit.com/Teardown/iPad+Pro+11-Inch+Teardown/115457
Apple has something in the same ballpark on size and is faster and uses less power. Why wouldn't they might want to use it? Why would Intel think they had some large leverage advantage here. They could look at iFixIt too and see what Apple has and look in their own labs and see what they have. If Apple hasn't asked for a reference board from this product area in a relatively long time ... it wouldn't take Columbo detective skills to figure out why.
[ Similar issue back late 2011 / early 2012 where there was a rumor here that Apple had asked for no Xeon E5 boards and hadn't engaged the OS "boot up" team at Intel. If Mac Pro hsd been on track that would be loopy. Yet in June 2012 Apple had about nothing new because they had asked for about nothing new. ]
Not really true. FCPX now versus 3 years ago has seen substantive updates ( about 10.2.3 is about 3 years back. FCPX change log . Need to scroll more than halfway down that page to get there. ) . LogicX roughly the same substantive progress. To try characterize the updates there as in roughly the same boat as the Mac Pro is mostly just misdirection.
The biggest Pro system in Apple's line is the MBP. Again over the last 3 years ... definitely seen substantive upgrades. Some folks may not like the, ( not enough variety of ports , no tinder inside options), but have had major attention applied. iMac got iMac Pro. Again not a speed bump upgrade. ( discounting because not a "box with slots" ) .
Is the Mac Pro "last in line" in terms of priority? yes. Is all of Apple's pro line up? No. Lower priority getting slower updates is yet another indicator that some overwhelming sweeping move off of Intel is probably not coming in a relatively short span of time. Apple doesn't move that fast over the whole line that fast anymore even when not changing platforms.
If the third iteration of Zen closes the gap to +/-3% percentage points to Intel , is more affordable than Intel , and Apple has worked out a 3rd party Thunderbolt controller solution ( or has sufficient certification with Intel's + AMD combo ). then could dump Intel in the desktop line up without much problem. Some driver and kernel work to do, but it would just be mostly a vendor swap for one that is executing smoothly over a couple of years versus one that isn't.
AMD probably doesn't have anything in the MacBook/Core "m" space either over next couple of years either. But they are significantly growing share in the Windows systems space. If they get to 20+% and Apple adds in 4-7% that is in the range of long term viable.
The problem is they haven't demonstrated that at all for the Xeon W class or high end desktop class CPU package (from AMD or Intel) . They are even further behind on desktop GPUs.
Apple has talked alot of smack about being faster than most of the Windows laptops shipped, but so are just about all of the current MacBook Pros. "Most" of the Windows laptops are sold at prices $400+ less expensive than Macs are.
If go to the $2K+ workstation market laptops ( especially those in the desktop replacement weight class) and the A12X isn't winning much at all. It isn't dominating in any way.
Apple CPUs have demonstrated nothing about being able to handle larger RAM capacities at all.
"... On iOS, 429.mcf was a problem case as the kernel memory allocator generally refuses to allocate the single large 1.8GB chunk that the program requires (even on the new 4GB iPhones). I’ve modified the benchmark to use only half the amount of arcs, thus roughly reducing the memory footprint to ~1GB. ..."
https://www.anandtech.com/show/13392/the-iphone-xs-xs-max-review-unveiling-the-silicon-secrets/4
Is there a MMU that handle large page tables and a variety of memory mappings... probably not ( because doesn't exist in the iOS devices that haven't reached double digits GBs let alone triple. digits. ).
Apple didn't particularly drop the "old" iPhone at all. It started ARM it is ARM now. The started to implement ARM themselves. A quite reasonable idea when they were using Samsung (a phone competitor at the conglomerate level) to implement ARM for them. The hard core fact is that they did NOT run off and leave a larger ecosystem that supported spreading the R&D costs out to multiple systems; not just folding the costs onto Apple. They didn't dump Samsung and then start over from scratch and invent their own instruction set.
ImaginationTech GPUs. Again they dumped them for a 'in-house' GPU over time but again a multiple year gradual shift. And they are still primarily running the ImagTech "front end" API on their GPU (just implemented by Apple). They didn't dump the GPU API completely and start over from scratch.
The run rates on leading edge iPhones is close to several 10's of million per year. The run rates on individual Mac models ( especially when get into the desktop classes) is an order of magnitude less (at least if not two when get into the iMac Pro and Mac Pro zone). Apple hasn't in-housed anything that is dramatically smaller volume at all.
If they are not putting tons of across the board effort into systems for which the basically buy a CPU "off the shelf" why would they be putting even MORE effort into pushing that component development task onto their plate. ( If the Mac is only being assigned 2-3 teams to development 6 product groups why would Apple put more people into components for those 6 than are are folks doing the overall systems ??????? If it is such a Scrooge McDuck operation on Mac system product resource assignment, how is that group of misers going to do some multi $100M chip component R&D spend ????? ) Or is it more likely the misers throw some "hand me down" chip over the fence at $100's of Million less in cost?
It isn't laughable that if they targeted late 2018 (back in 2017) and things slid into early-mid 2019. Just about everybody who has coupled to Intel's CPU products over last 2-3 years has had 1-2 quarter rollout slides. When Apple does release something they will have waited from some set of components they picked to get to volume production.
That the CPU/GPU were the cause of the "pause button" being held down from 2014-2017. No. It wasn't. That's far more so probably do to the fact that it really wasn't a priority for Apple or a significant chunk of their users ( witness "are you happy with your 5,1 " threads that have appeared recently.).
No. The first iteration of Itanium was shipped in 2001 ( the work started in 1999).
https://en.wikipedia.org/wiki/Itanium#Itanium_(Merced):_2001
There was NO complete x86-64 in 2001 in the form of widely deployed binary code. The first x86-64 Opteron didn't ship until 2003. So no, they didn't have a TARDIS powered execution mode to execute code that didn't exist yet. Legacy ( older ) code was the focus.
"... Intel has announced that they will be releasing a software emulation product to allow 32-bit x86 apps to run on Itanium Processors ..."
https://hardware.slashdot.org/story/03/04/27/2313248/intels-itanium-will-get-x86-emulation
one of the links off the slashdot article should point to
"Thirty-two bit support across 64-bit systems is important to Red Hat customers, and we are exploring ways to address this across the various architectures we support," Brian Stevens, vice president of operating system development, said in a statement Friday.
https://www.cnet.com/news/red-hat-warms-to-itanium-booster-plan/
"... Itanium chips currently include circuitry that lets them run the 32-bit software of "IA-32" processors such as Xeon or Pentium. But that circuitry's performance has been so poor that not even Intel advocates its use. ...
... Intel will keep the hardware-based IA-32 support at least through the Itanium II 9M model due to arrive in 2004, Grimes said. She declined to say whether the company would rely solely on the Execution Layer software after that, but analysts believe that move is likely. ... "
https://www.cnet.com/news/intel-plans-itanium-course-correction/
The emulation overhead for the 2001 model was far more aimed at PA-RISC ( which was 64 bits, but not particularly the 64 bits you are alluding to.). It was mainly being aimed at pre-2000 code and trying to consolidate that workload onto newer Itanium systems. This is one of the situation where Steve Jobs alluded to where "sometimes IT folks are wrong ".
So there really was an effort to run 32-bit code on a VLIW skewed machine. Eventually it got removed.
".. .Itanium processors natively supported the x86 instruction sets, but they stripped the x86 hardware emulation feature from the processor due to its bad performance. "
https://wiki.osdev.org/IA-64
That 200-2003 x86 emulation code support transistor budget and allocated simulator time ... a waste of time and effort. ( maybe a minor precursor to x86->RISC front end decode issues , but mostly a really bad mismatch of designs).
It was a relatively obvious bad idea but x86 was dogma to some extent at Intel and some customers back then ( well now too. And not surprising had the Larrabee x86 GPU escapade in between either ). Itanium had a chance to be a PA-RISC, MIPS, Alpha, SPARC , and Power/Z killer. If they had purely focused on that it would have had a better chance of running the table on all five (about 3.5 out of 5 isn't so bad). x86 was a side show. Where Power survived was really more so about Linux on Power. Getting the software off the proprietary Unix and consolidated onto bigger iron helped. There was no "shortcut" with emulators.
if Itanium wold have moved quicker to get settle into the space where eventually the Xeon E7 and SP and Epyc sit now it probably could have developed more inertia to be harder to displace. Instead of x86 emulation, they could have worked on some of the "out of order" stuff (and perhaps also branch prediction) that got deferred to far too late to help defend the space. ( they needed to pull some of the dynamic stuff that the compilers were always going to be "moonshot" hard problems and push that back into the processor. )
For macOS the legacy code inertia will probably turn up again. x86 code on ARM even with a JIT-compiler to drive emulation is going to be slower. If all Apple can get to is "just about" even parity then the emulation overhead makes it a negative. low end laptops they have "horsepower gap" to spare, so it would work well. In mid-high end desktop space, they largely don't so it probably won't.
Neither Intel or AMD are trying to do what Intel was trying to do with Atom about 3-5 years ago. That is a defensive barrier that ARM essentially cracked. But as AMD and Intel retreat a bit to defend a smaller space that is probably going to make it harder for ARM to displace them. ( ARM server stuff has wondered in the swamp for many years. Its getting traction at the edges there now too, but it has taken a long while. ) AMD has jumped in with Intel to fend off ARM in the Chromebook space and it is working relatively well so far ( App deliver from the web services evens the playing field on both 'sides' ).
As long as AMD and/or Intel create CPUs that are focused on defending the space the iMac/Mac Pro are in, they is more than decent change they can keep that spot x86_64 based.
https://superuser.com/questions/276994/difference-between-ia64-and-x64
"IA64 refers to the 64 bit Itanium architecture while x64 is the 64 bit extension to the x86 architecture"
https://en.wikipedia.org/wiki/IA-64
"IA-64 (also called Intel Itanium architecture) is the instruction set architecture (ISA) of the Itanium family of 64-bit Intel microprocessors"
You’re right @deconstruct60; intel isn’t the only x86 vendor. But Apple
Still hasn’t put in any other chips. My position is that they are looking elsewhere in a different direction. Apple could have kept a PPC architecture and x86 since it was ‘living a double life” since 10.1, but didn’t-sure they had supported universal binaries for a time but not longer than the last time the Mac Pro was updated. Are you really saying there isn’t any possible chip that could be better then what is currently in the machines that is out? Regardless of supplier-intel/amp?
The 6,1 was a f*** up, and Apple really hasn’t had anything they could do? Nothing to mitigate the situation? The iMac Pro? (Where the speculation is there was enough blow back that they went back to the drawing board for a pro desktop?)
C’mon this is a 3-6 year flustercluck at this point.
[doublepost=1552439365][/doublepost]Is there anything stoping Apple from making their own x86 class chips?
Still hasn’t put in any other chips. My position is that they are looking elsewhere in a different direction. Apple could have kept a PPC architecture and x86 since it was ‘living a double life” since 10.1, but didn’t-sure they had supported universal binaries for a time but not longer than the last time the Mac Pro was updated. Are you really saying there isn’t any possible chip that could be better then what is currently in the machines that is out? Regardless of supplier-intel/amp?
The 6,1 was a f*** up, and Apple really hasn’t had anything they could do? Nothing to mitigate the situation? The iMac Pro? (Where the speculation is there was enough blow back that they went back to the drawing board for a pro desktop?)
C’mon this is a 3-6 year flustercluck at this point.
[doublepost=1552439365][/doublepost]Is there anything stoping Apple from making their own x86 class chips?
I stand corrected.
The Mac Pro round-table coincided with the time that Apple would start showing around the new iMac Pro to select customers in order to determine sales estimates before release. The feedback they got must have been pretty bad because they did a 180° on the Mac Pro just after that. It is not normal for Apple to pre-announce a product as far in advance as they have with the Mac Pro. Apple's actions demonstrates their embarrassment. In order for Apple to make a successful Mac Pro product they need to make a machine that they can sell consistently after release.
The iPad Pro couldn't last minutes under a heavy workload - if anyone was silly enough to try and be productive in iOS and its ancient input options , using any demanding software .
Selective, synthetic benchmarks, cherry picked for a brief media event, don't turn a gadget into a workstation .
Nor does that gadget become a giant killer by simply putting some of them into a bigger box .
AFAIK, iOS processes get killed if they spike the CPU for more than ~8 seconds - so a question - is the A(x) chip itself large enough to get sufficient surface area contact with a cooling structure that it could bleed sufficient heat to survive running continuously at 100% for minutes, hours, days etc?
A huge contributing part of that is that from around 2005-2017 AMD was screwing up their x86 product line up at least as bad as Intel is doing now (at some points along that time line was substantively worse ) . They were not a competitive options. 3-4 years it was certainly prudent for Apple to do preliminary exploration of a "big bang' option. In the current contents, (and where they have bigger modem drama) not as much.
Back when TSMC was less predictable and behind the edge of Samsung ... Apple used Samsung fabs. Now TSMC is executing substantially better Apple is leaning contracts their way. If Samsung could jump substantive out in front of TSMC with a highly reliable process, Apple would probably jump back.
If AMD stays on track they are viable option. Apple 7nm SoC isn't going to particularly out "nanometer" AMD's 7nm process built in the practically the same TSMC fab. Money puts toward AMD CPUs would go toward paying down the cost of the same close to double digit billion dollar fab that Apple other stuff is in.
The recent "foul up , bleeps , and blunder" on process fabrication aside, Intel also jumped through many of the hoops Apple put forward. Thinner CPU packages, higher performance iGPU (more die allocation for GPU), joint work on Thunderbolt , OpenCL support progression , etc. On the GPU (or CPU) front AMD really hasn't jumped through hoops as well over that 10+ year track record. AMD has done a few, but not as broad a scope as Intel. [ Very similar issue with Nvidia. ]
If AMD has jumped through all the hoops Apple laid out 2-3 years ago for 2020 product better than Intel has then there is a decent chance Apple will change vendors.
I think niether Intel nor AMD probably cut it for the MacBook product but for the rest there is a shot. Much better shots as go 'up' the Mac Pro line up in terms of performance floor.
Yeah they could have ponied up a large chunk of money to have IBM (or Motorola ) make them a suitable broad spectrum line up. They didn't. But that's primarily the point ... they didn't want to pony up a large chunk of money just for narrow Mac product usage. Trying to make it billion dollar 'rain' on Mac products components wasn't the objective then or now for all the objective evidence.
Apple didn't treat fat binaries as a long term tool, but far more so as a transition tool. They chucked PPC after a transition window primarily because probably for two factors. One, they had transitioned the hole line up for more than a couple of iterations. So transition time was "over". Second, when they chucked PPC that opened up a slot for whatever might turn up in the future. In both cases, Apple is primarily focused on where things are going to , not making people more comfortable with living in the increasingly distant past.
If things going forward is far more split classic PC market ( Windows + Mac) then being on a split path would be the future.
I'm saying there is nobody with an open roadmap that is targeting the heart of the Windows PC space so it is highly unlikely there is another vendor to jump to because Intel/AMD offerings. Apple isn't a particularly creditable vendor to fill that space. If nobody was moving toward that space 2-3 ago then there won't be much of anything there for the next 2-3 years either. Supplying CPUs into markets typically have multiple year lead times (at least .... ARM for servers has been ramping even longer than that). Qualcomm has 'something' lined up for "Always on , Always connect" Windows laptops, but that is largely been a "hand me down" offering ( not to unlike Apple.). It is probably going to do next to nothing in the desktop space.
If Qualcomm can get 10% of Windows market to peel off that would be a segment that is larger than the entire Mac market on one close product on ARM implementation. 80% of the x86 Windows market would still be 8 times bigger than Mac. Both highly viable situations. Splitting 10% off the Mac market's (rounded up) 10% is just 1%. That isn't so viable by its lonesome (outside of almost completely leaning on some iOS product for volume). the 9% thrown in with the 80% Windows market completely is.
What Apple is likely to do with the ARM line up is just as likely to be that it will be used to prop up volumes for the iOS products (eg.. iPad Pro) more than to does something entirely forked to make the Macs "special" again.
Apple went "back to the drawing board" for iMac Pro? Highly likely not. Most the venom spewed at the iMac Pro by hard core old school Mac Pro users was extremely likely spewed at the Mac Pro 2013 (6,1). So it likely wasn't anything new or revolutionary. The folks who primarily wanted it to be a "box with slots" would be harping on form rather than function for both different forms would. So it would have been much of the same stuff. So back to the drawing board? Probably Not. Had already heard that before doing the iMac Pro.
There are some edge cases that the iMac Pro leaves off dual GPUs, more internal storage, single network socket, etc. , but how much wouldn't be clear until they got some feedback. There were changes in a literal desktop solution between the 6,1 and iMac Pro. For a subset of the 6,1 target audience the iMac Pro is a correction. And because many of those folks actually bough something in the last 3-4 years they get priority.
If Apple is going to put these into somewhat overlapping space they probably would want to measure twice before cutting for the next Mac Pro.
How Apple has gotten to the molasses product update state on the Mac product line in general is problem. It isn't all 100% Apple though. Slower updates from customers is a contributing factor. As the number of folks buying Mac Pro every years goes down, the priority is probably going to go down too at Apple. It is a basic feedback loop that Jobs set up and is firmly woven into most of the basic approach Apple follows.
Not having a license and the intellectual property to do so is a sure show stopper. It would be cheaper for Apple to buy a licence for ARM 'N1/E1 platform and just try to tweak those in 3-4 iterations than to jump into x86. Given apple is the likely only consumer/customer, going x86 at 100% in-house doesn't make much sense. It is just going to be cheaper to just buy from the vendors there.
[ And no, just buying an x86 license isn't particularly viable. ]
Just about the same boat as how come someone doesn't make a macOS exact clone.
So, if the Mac Pro is expected to arrive later this years, why are we barely hearing anything about it? Where are renderers, concept art, where are guesses from Kuo, Gurman...why is it all so quiet??? If it was the all new iPhone getting released after years of neglect, you would have sites like this going crazy.
People say that Apple does not care about Mac. And that might be true, but guess what else I think is true, people do not care about Mac either. It is just a handful of diehard fans, and that's it.
People say that Apple does not care about Mac. And that might be true, but guess what else I think is true, people do not care about Mac either. It is just a handful of diehard fans, and that's it.
The Mac Pro round-table coincided with a price drop for the Mac Pro 2013 model. The macrumors count up clock was about about to shoot past some milestone marker. ( and was reset for a while with price change. ..... but buyers guide undid that.
"... Apple's radically redesigned Mac Pro hasn't been updated since December 19, 2013 aside from an April 2017 price cut, .." ). Still on track for busting through the 2000 day mark before WWDC.
The meeting was also a big clue for an upcoming iMac Pro, but they were doing direct Mac Pro juggling there also. The iMac Pro was not intended to be the "missing" Mac Pro that the Mac Pro 2013 was not.
Errr, no. The price cut and sticking the current Mac Pro into "placeholder" status was not a huge shift. If the Mac Pro was about to be 100% obsoleted by the iMac Pro in a matter of months why cut the price?
The iMac Pro was eventually placed in a pricing slot above the Mac Pro, Again a symbolic that it is a different product ( these two aren't the same thing in Apple's mindset. Nor should be in customer's. )
The feedback from the folks who were entrenched in not wanting an iMac Pro was probably pretty bad, but highly doubtful that with folks who are not predisposed to be viscerally against the iMac form factor that it was huge failure. The iMac Pro continues to sell at a relatively good clip. In no way is it some huge failure sales wise so as long as Apple was talking to an accurate customer group it extremely was not probably "pretty bad". Most folks don't go out and buy "pretty bad" stuff.
What Apple did essentially say in that meeting is that had focused their efforts on a more "pro" iMac. When Apple only has limited amount of focus time for the entire Mac product line that far more likely means they just were not working on a Mac Pro ( lower down in the priority queue). there is little objective evidence that Apple has done a 180 here at all. The Mac Pro is probably on a lower priority queue.
If are measuring from 2011-12 to 2017 it isn't normal for Apple to be 5-6 years late with a product they supposedly care about. That horse kind of left the barn. Apple's huge constraint is that they can't talk about future products. They can talk about current products. The meeting was more about the current Mac Pro ( which just had a price change) than about the future one. The hope was that folks would fill in some of the blanks on the future product from the "didn't turn out so well" aspects of the current one.
that was probably helpful at first but at this point things are going way off into the weeds of "Lego" Mac Pro systems in just Whacky races stuff.
Embarrassed with substantive higher revenues of the Mac product line up?
Embarrassed with substantive higher unit volume than 2-4 years ago of the Mac product line up?
Embarrassed with substantive services revenue growth leveraged off the Mac produt line up over last 2-3 years?
Probably not. Perhaps concerned that asking folks to circle the airport for another two years is big "ask". (even for folks who mostly wanted to just 'sit' on their the product for >6 years. ).
Apple had more than just the Mac Pro in comatose status in April 2017. The "more than 1000 days" mantra on the Mac Pro was far more just a lead in to several "Apple is abandoning the Mac" discussions that were popping up. That's way the talk was based in a machine shop. "We are grinding away at real metal in here ... we haven't stopped." was a major part of the story they were presenting.
iMac Pro 2017 , Mini 2018 , Mac Pro 2019 .... see haven't completely abandoned at all kind of story. (perhaps can't chew gum and walk at the same time, but can slowly walk. ). About zero evidence though that they have changed from the "walk slow" plan.
If "consistently" is an implication of yearly, "big news" upgrades then probably not. Not for the Mac Pro or most of the rest of the Mac line up.
[doublepost=1552502626][/doublepost]
The likelihood of leaks is proportional to the number of folks working on it.
The supply change on the iPhones is huge. Lots of suppliers. Lots of workers at those suppliers. All that leads to much higher likelihood will run into blabbermouth and/or someone out to make some money.
The Mac Pro probably has a very narrow set of suppliers and relatively minuscule internal Apple assigned to it. Apple Security being able to track down the leak is much more likely. For many of those folks it isn't worth being fired at all. Similar with customers shown early stuff NDA.... not worth pissing that kind of Apple access down the drain.
The folks covering leaks from the financial analyst perspective are slightly different. The Mac Pro about zero impact on Apple's balance sheet. Since it doesn't matter on that dimension they don't look all that hard. It is super easy to miss what you are not looking for.
Kuo .... finance ( and has recently said something. so perhaps may have stumbled across that. Not hard since Apple openly declared in April 2018 that it was a 2019 product. )
Gurman ... software leaks are better than the hardware leaks.
Folks who use tools wouldn't necessarily see much. Mac Pro 2013 the NDA group didn't grow to leaky size until 2-3 months before the public preview.
Apple cares about the Mac product line. It isn't more important than the iOS line up (tie breakers are going to fall iOS way). The Mac Pro isn't the Mac product line. The Mac Pro is not strategic ( there are folks who arm flap that it is strategic but it isn't. It is basically in the "nice to have" product status at this point. Neither strategic or essential. )
Flagship as a euphemism for expensive .... is just that ... a euphemism; not literal.
This is definitely not going to happen.
The nutty professor YouTube clown illustrated a similar idea where he tried to separate as many parts of a computer into individual Mac mini shaped boxes as he could.
Apple will not pursue this kind of thing for many reasons, including:
- It would look fussy and unattractive. Apple design elegant, minimalist products. This would also be much bigger than necessary. So much added complexity so that users might add or a change a module every few years, or possibly never.
- It relies on the end user to put it together. Apple are all about one or two plugs, and you’re ready to go, not a Lego set. End users also mess this stuff up, however obvious it might seem.
- It’s massively environmentally wasteful. Apart from the additional materials and manufacturing required for each little box, there’d be additional connectors and tiny little fans and plugs galore. Also it’s a bad look when people start ditching whole modules when a new standard comes along. Maybe some of your listed modules aren’t needed by everyone, but most are.
- Warranty claims would become confusing if modules were purchased from a variety of vendors over a time period. With loads of extra points of failure to boot. What if people start cobbling together systems made up of other people’s cast offs? Apple will inevitably be drawn into unnecessary bad publicity when those systems fail.
- It’s confusing. Non-nerds may well not be sure what they need, even if they require a high-powered system. Plenty of people in my field of post production are amazing creatives, with literally no idea what specs are in their machine. Seriously.
- Pricing and tiers are gone. Apple generally have two or three choices and prices per product. So you can gauge the cost of things at a glance. If you know your stuff you can delve into the BTO section, but the tiers are easy to follow. Having to to suddenly find you need more PSU modules as you configure and watching the price jump up? Not a good look.
- Supply chain nightmares. Apple are expert at supply chain management. Having to package, store, track and ship one product in lots of separate boxes would lead to more mistakes and more cost. For what gain?
TL;DR... the module stack appeals neither to nerds who want to customise their system, nor to people who don’t care about the specifics but just want ‘the powerful one’.
Apple won’t make the module stack.
Last edited:
I don't think this is true at all. My bet is that since the Mac Pro is manufactured in the USA by Apple, it's much less likely to get leaked. The reason iPhones get leaked so often is because they're manufactured by a variety of third party companies all across the world.