So just reading the "tea leaves", where does that leave the mMP?
With alot more performance headroom than the iMac Pro. Just 3-5 pissing matches folks folks on both side ( Apple and "generic box container with slots" ) need to give up on.
It seems from Apples POV, the iMac Pro is the "perfect" desktop workstation for 90% of that market (that would use a Mac).
While the Mac Pro is in the "desktop" category Apple needs to step back from taking that literal. The Mac Pro ( and Power Mac, Quadra 950, ) were really desk-side computers. Bigger and more bulky that what will comfortably fit on most desks.
So yeah as a desktop computer the iMac Pro is better. Lower actual desktop footprint than the Mac pro will get. Pushing the Mac Pro into a footprint smaller than the Mac Mini was a contributing factor to painting themselves into a corner.
Most of the market the iMP is not serving is either 1) people who like to use their own displays (i.e. want nothing to do with a glossy built-in display no matter how many other external displays they can hook up to it), but Apple doesn't care about that (as any computer they offer without a built-in display languishes for years without updates),
1. Apple cares somewhat since they also said they were going to jump back into the "pro" display market. While technically, that will likely be a display docking station, there are probably expectations that Apple will sell some of those to Mac Pro users. But yes they probably do expect the majority of the folks to go for 3rd party options. I extremely doubt this is some "secret" that Apple doesn't already know.
2. Those 3rd party options will extremely likely have DisplayPort sockets. Thunderbolt v3 ports work even better with DisplayPort with the new (and available for a new Mac Pro later in 2018 ) Titan Ridge controllers (
https://www.anandtech.com/show/12228/intel-titan-ridge-thunderbolt-3 ) which pass through DP v1.3-1.4 just fine.
four TBv3 (titan ridge ) ports plus two HDMI 2.0b (maybe 2.1 if a new GPU can drive it) ports makes for absolutely zero disconnect between a new Mac Pro and 3rd party monitors. The number of 3rd party, "pro" monitors that have neither DP nor HDMI is too small to worry about.
There is no 'disconnection' here at all with a updated Mac Pro with 4-6 Thunderbolt v3 sockets and third party monitors.
[ The mini is a somewhat different market space than the Mac Pro is. The primary purpose of the Mac Mini was to be the most affordable Mac. There is lots of market pressure on Apple to push the MBA into that role ( or rename name it MacBook) and pushed in to cover vast majority of Mini's price range. ( $300-500 Chromebooks are eating Apple alive. MS has Windows S . ) . If Apple keeps the MBA , MscBook , and multiple variantions of the MBP then the Mini may go because there are no resources to be allocated to it. Desktop vs Laptop trends, the latter has been winning (scale, number, and revenue ) for about a decade now. The Mini was always more a "headless" MB/MBP than anything else ( it wasn't the anti-iMac many want to cast it as; doubtful Apple views it as such). ]
or 2) users who like to do their own internal upgrades, again, a market Apple doesn't care about or have any interest in encouraging.
This is part where some users are clinging to "form over function". The Mac boot and GPU card environment are driven by work and contracts made by Apple. Generic, legacy PCI-e slot standard cards don't necessarily work in a Mac and certainly don't integrate with Thunderbolt all that well. The boot screen GPU something that Apple should "drive". The T2 drive points to Apple taking control of the standard boot drive also ( It is a company that makes SSD controllers so why would they stay out of that market? )
Apple does need to show some more flexibility with the new Mac Pro when it comes to areas outside of booting and the standard OS/Apps/User home folder storage. folks with lots of data don't (and in many cases probably shouldn't ) use one storage volume for everything. So while there Apple sockets for GPU and secure boot SSD the sockets for a 2nd GPU , additional SSD storage , or perhaps bulk data store (HDD) would make more sense to be 3rd party.
Two M.2 slots , one-two standard PCI-e slots (e.g., one full elect/mechanical x16 and perhaps another x16 mechanical / x4 electrical ) , (and perhaps one-two HDD drive slots. Although if rest of Mac product line is dropping them then perhaps not. ).
Both sides wanting to completely exclude the other is dubious. There is a sizable market in the middle.
The market for *Mac* computers with more power than the iMac Pro is tiny... not the kind of sales figures Apple's interested in.
There are multiple dimensions to "power". ( lots of threads in this forum turn into AMD vs Nvidia pissing matches but that is relatively myopic in insight. )
It isn't necessarily more "GPU" horsepower. Some folks are stuck with sunk costs. For example a shop that sunk $15K in to some Avid Pro Tools rig that is coupled to a PCI-e card. To some extent, to those users that PCI-e is about a $15K investment. So if buying a new $2-4K box to put the card into then putting priority on that box is the tail wagging the dog.
8-10K HDR multiple stream video capture could be tough to do with an iMac Pro. ( a x16 slot would be handy)
Nvidia CUDA is another sunk cost pit, that Apple isn't going to rationally move folks out of with the iMac Pro.
It also isn't necessarily more "CPU" power than iMac Pro. The iMPro has the same sub 500W saddle the MP 2013 had. The iMPro's GPU is down clocked. There are a sizable number of folks who want that nominal clock rate otherwise AMD wouldn't have built it that way.
The Mac Pro could use the same CPU and have 8 DIMM slots. It isn't more "power" computationally it is more storage.
There is very substantive I/O bandwidth that the iMac Pro leaves sitting on the table.
The Mac Pro doesn't have to be entry priced higher than the iMac Pro (e.g., doesn't have to start at 8 cores) . If Apple artificially pushing the price higher then yes it will be death spirally small market, but there is no need to do that. It also still could be priced higher than where the bulk of the "xMac" crowd want. So they will still moan and groan.
So is the mMP ultimately mostly for "show"... to keep the tiny niche happy (along with the press that likes to focus on these things) because it has outsized influence in the success of Apple?
No. I think there were some folks inside and outside of Apple who thought that. The "car show prototype hype car" that is the "flagship". Roll one out every 3-4 years as tech demonstrator.
Three tactical factors that Apple needs.
First, if they choose 'wrong' on iMac Pro balance they need something so that can gap fill for a couple of years while they adjust. [ I suspect not going to see iMac Pro or Mac Pro do yearly major upgrades. The cycle could be longer than 2 years. If there is some unexpected "inflection point" card that comes along then users can just throw it into the secondary PCI-e slot. ]. What hurt Apple was not so much that they made some questionable calls on the MP 2013. It was more so that they had nothing to respond (and communicate) with for years when they did. That is a mistake they probably should try to avoid. ( can keep the "no talk in detail about future products" policy but they have to
do something on a regular basis for that to work. )
Second, If they choose right on iMac Pro then Mac Pro with same core stuff with more thermal headroom ( can ride the economies of scale of the common major parts to keep "good enough" margins to continue. )
[ very similar to what Mac Mini used to do riding in the volume laptop major parts bow wave. ]. Honestly the iMac Pro isn't likely going to have sky high volumes either. ( it is likely going to be in the some 100K/year run rate. ). Two 50K product lines with high major component overlap could effectively be a 100K/year effort. [ The question Apple has to get a grip with is whether this upper workstation market is segmented enough so that these two are largely two hard core camps and don't engage in a significant degree of fratricide. ]
Third, this keeps somewhat of a lid on the hackintosh market. There will still be disaffected folks, but as long as that remains a sub 1% of the Mac market then the detente can stay in place. Also, used Mac Pros don't do anything for Apple's bottom line, but they fill part of the "xMac" market. Even if the iMac Pro + Mac Pro amount to 2-3% of the Mac market that is better than letting a large fraction of that flip to the hackintoush crowd.
It is also a bone to keep Nvidia (or AMD) in the GPU "bake offs" that Apple runs ( or invite in Intel's discrete GPU if that takes off). I doubt eGPUs by themselves will cut it in terms of volume.
Strategically, even if Mac Pro only last another 4-7 years at point the technology the iMac Pro would have at that later date would cover even more the "fast enough" for an even larger crowd. The new Mac Pro could sever as a better gateway to expanding iMac Pro sales even if it disappears later.
It seems that much of the challenge in guessing at what the mMP will be is in figuring out what Apple's "true" motivation for developing the mMP is... and even more challenging, has that motivation shifted at all since they (internally or publicly) committed to a mMP?
There is zero evidence that the motivation has shifted at all. I think there are substantive numbers of folks who took the stories coming out of the meeting from last year and built a "we won, we are getting entirely the old Mac Pro back" story out of it that wasn't there in the first place.
Instead of guess can take a look a the iMac Pro press release ( which completely aligned with what got said in the meeting)
".. In addition to the new iMac Pro, Apple is working on a completely redesigned, next-generation Mac Pro architected for pro customers who need the highest performance, high-throughput system in a modular, upgradeable design, as well as a new high-end pro display. .."
https://www.apple.com/newsroom/2017/12/imac-pro-the-most-powerful-mac-ever-available-today/
Apple talked to folks and looked at where they previously sold Mac Pro class machines they asked for that set of stuff. Apple looked at that functionality ( not form ) and thought it was a big enough market that enough folks would buy them. That doesn't say "we are going to build a HP z8 clone" . The iMac Pro has a 500W limiter and leaves a chunk of the PCI-e v3 lanes out of the CPU completely unused. So it doesn't meet those exact criteria. Take off the screen , allow DIMM upgrades, bump the limit up to 800-900w, allow two GPUs in some BTO configs, and use the other PCI-e v3 lanes and you'd have highest performance of a Mac and higher throughput of any Mac. And Apple would like for you to buy their display (docking station) too, but the box will suffice.
Motivation as in why has it taken Apple an eon to get a new product out. That part is questionable as to how transparent Apple has been on that front.
