We're talking about the company that believes everyone should be happy with Thunderbolt GPUs right?
Some other alternative universe?
1. If Apple is so eGPU only why did they bump the MBP 15" to Vega 20 and the iMac to Vega 48 ? A second or third ( or fourth ) GPU Apple may punt on but not keeping the system's basic GPU up to latest available that fits the criteria. Apple doesn't have a deep track record on that. ( Apple's criteria likely includes cost and other factors than simiply based upon the latest GPU tech porn (and/or fanboy ) benchmark. )
2. If it was so extremely strategic to them why do they make zero eEPUs. None. Super duper essential to the product but don't make the items. Does that sound like Apple? No, it doesn't.
Are eGPUs a 'nice to have' augment to Mac systems. Yes. Apple leaving that as a whole 3rd party opportunity. Totally inside the Apple playbook. 'Happily use a MBP as a desktop replacement by augmenting it in the stationary space with we desktop eGPU" ? Sure. But the MBP will work well detached also on the move.
3. When an eGPU is used to drive visual display, the standard Apple recommendation is to hook the monitor to the external device. Is Apple really going to encourage folks to hook their monitor so something other than a Mac Pro? Probably not. The Mini still has a HDMI port. That is demonstrative that they are taking some extra steps to get that monitor hooked to the Mini itself (not some docking station. ). The notion that our Mac system is not something you'd want to hook a monitor to is far from being a primary design objective. Some folks may want to in some contexts but that isn't a primary, "baked into the design" objective. And there is zero indication that is seperate track they'd want to put the Mac Pro on.
For the Mac Pro there is likely some number of GPUs that is greater than one ( > 1 ) that they will point to eGPU as being the option. If the start at one then that will be controversial ( zero empty slots and no way to add ... putting into same space on that dimension as iMac , iMac Pro , and mini). if they have it start at " > 2" ( i.e., one empty slot). but only after some relatively expensive gyrations of extra equipment.... that would be very short sighted. The future trend lines on what a slot would probably need to cover ( PCI-e v4 -> v5 -> v?) is yet another "painted into a corner case" waiting to happen.
Sure, yes, that would be super complicated in a workstation and unnecessary but uhhhh... again... We're talking about Apple.
Apple who primarily using the embedded GPUs they have worked with to compose their list of "Approved" candidates to be placed in an eGPU. Apple also knows that driver and boot support needs (or should) be there. eGPU hardware in and of itself isn't a panacea.
Apple's complexity of design typically has to do with integration ; not disintegration. The primarily objective typically is to have a relatively whole, working system. ( pull from the box , plug it in , plug in some minor set of peripherals (keyboard , mouse) , and it basically just works on core functionality. )
It does not seem out of character for Apple at all to say "all GPUs are external." That sounds exactly like Apple.
No it doesn't. They don't do it with any other product. iOS devices the primary "we do a great job" is how they have
integrated the CPU and GPU onto the same package. Mac Mobile devices ... same thing: embedded best GPU they can embedded with constraints.
Apple doing a GPU-less ARM ? Does that sound like Apple? Not on their track record at all.
The trend and focus of Apple is toward products that have a screen/display integrated..... So integrate the screnn but leave the GPU outside? Doesn't sound like Apple at all.
The Mini and the Mac Pro deviate from the integrated screen track but they are the minority of the line up. ( and part of the line up that Apple has "kicked the can" on the most over the last 5-6 years). Most likely Apple is probably going to try to keep those two outliers closer to what the others are doing; not farther away. ( Massive integration of Mac Pro 2013 being extremely representative of that. )
Again, we're talking about SATA sleds and cost savings for customers... did I wander into the wrong forum? This is Apple we're talking about.
Apple who has had the XR on trade in discount since launch? That didn't drop the MBA when the retina MBA came out? (e.g., now keeps older models around longer in the Mac space. ) Apple has tested the price elasticity of its customers to about as far they can. It isn't primarily cost saving for customers it is cost saving versus alternative options. The folks who 'passed' on Mac Pro 2013 and iMac Pro and are still circling airport for a new Mac Pro mostly do have very realistic alternatives. Folks at Apple would have to be drinking gallons of Cupertino kool-aid to not know they have a net "outflow" of switchers in this product space.
This is not about Apple competing solely on equal costs (and discount sales). It is far more so about how much "Apple Tax" they can add on top before it gets counter productive. The SSD $/GB prices probably will be out of whack with the then current market reality. All of your data on one single drive has problems in the wider workstation space. If they drive the costs of adding SATA sleds up past what a DAS/NAS/SAN is then sales will collapse on the augment and if it sells badly then Apple will tend to abandon that space. ( yet another Rip van Winkle development cycle isn't going to help this product long term. )
There's nothing stopping them from building some higher bandwidth proprietary version of Thunderbolt and basically treating every component in the stack as a short hop Thunderbolt device.
There is lots of things stopping them including "return on investment". Some quirky , proprietary interface just for the Mac Pro is too low volume. Apple composing some wierdo connector for even just the desktop probably wouldn't work either. Not enough scale to get a viable infrastructure.
A custom connection that combines some established connection ( a variation on SATA and power that better fit a "snap on" constraints ) perhaps. But remember Apple went to Intel with LightPeak to get a partner. Apple now going partnerless ....... that would be a huge shift. No partners would be proprietary hole forever. For low volume and no hope of some very high growth phase where is the return?
Is it probably a bad idea? Yes. Does that sound exactly like something Apple would do? Does to me.
Bad ideas that loose alot of money doesn't sound like Apple to me. Apple has done stuff that has been corner case and put more money in their pockets. But forked off of Thunderbolt to do something odd primarily just for the Mac Pro ? That is a huge leap.
I would personally hope for the backplane idea over stackable. But again, stackable sounds like exactly the sort of thing Ive would come up with. Backplane isn't "Apple" enough.
Stackable sound more like some folks outside of Apple hysteria extrapolating on some narrow aspect of the Mac Mini intro demo.
Apple putting a more limited constraint than volume of a Mac Pro 2012 case into order to use "recycled aluminum" and/or cut the carbon footprint of the new product ... Yes. Empty space for drives that 30-45% of the users probably won't use ... yeah that probably could get axed and perhaps an option put in for the folks that were going to completely fill it.
A new Mac Pro with a T-series chip and a several other Apple only touch points is "Apple" enough and are more focused on 'value add' that only Apple can provide. An SSD drives only would be "Apple" enough ( since APFS and most of the system software is focused on the SSD future).
If there is just one empty PCI-e slot that too would be "Apple enough" since macOS driver support would need to be there too. It doesn't make it "Apple enough" if 80+% of the Mac Pro user space is using at least one PCI-e slot and Apple pulls that out. That's is just taking the value add opportunity away. It highly likely won't help them sell more product. And even less product if some unforeseen addition comes along (e.g., some xyz compute or i/o ) card comes along that need higher bandwidth (and/or lower latency). Apple gets caught sleeping at the wheel again on design refresh. Also not particularly "Apple enough" is just selling a highly fraticide product that almost completely overlaps another product being sold. ( how going to sell Mac Pro alongside an iMac Pro is both held to pragmatically the same set of constraints. )
[doublepost=1557506018][/doublepost]
Where was the exhibit hall , showroom floor were loads of people wander around gawking at products. None.
Products is the plural of product. ;-) Still none. [ tap dancing and throwing corner cases, like the news media showcase after a keynote and this one time stunt , is beside the point. ]
Additionally, it is doubtful that an early verification "mock up" in a plastic case is going to significantly help them this time. It isn't gap from 2010=2012 ( 1-3) gap they are trying to paper over here with a non shipping product. It is a 2010-2013 ( 6-9 ) year gap. Patience for most and severe detachment from the their upgrade schedule wants/needs went out of sync over a year ago. They aren't going to get a "free pass" on some 'can't innovate my ass " spin. Nothing in 6+ years is complete non innovation.
This product is really a turd in the punch bowl at this point even if they did it in a way to sell "good enough" to survive. WWDC is highly likely the wrong place for the intro. If it was going to ship that day perhaps but essentially yet more "dog ate my homework and even STILL not done yet and still not have a firm date because the kinks aren't worked out yet " doesn't fit with WWDC augment with actual product.
[doublepost=1557506754][/doublepost]
Of course the Mac Pro will be unveiled at the not MacWorld not San Francisco event in June and the MacBook Pro refresh featuring 9th gen Intel processors will debut unceremoniously just before.
If there are new laptops those would be far more likely candidates for a limited stop on stage during the keynote. Especially, if had incrementally improved keyboard along with the processors. That's is a no brainer substitute for a "not going to ship for many months and still not done after 6 years" product.
Apple has stated that the Mac Pro will be a 2019 product and what better time than One more thing... for Apple to introduce the most anticipated product of the year.
Apple already has "one more thing". WWDC has to cover macOS , iOS , watchOS , and tvOS. They already have
four operating systems to cover on stage. If the iPad is slightly forked from iPhone iOS then that's
five ! The keynotes drags on way too long as it is. .... adding long footnote at the end that Apple managed to screw up and it will take even longer to get the Mac Pro out is
not an uptick to end on ( 'we screwed up isn't a 'uptick'. ) .
For this WWDC specifically, Apple is likely going to spend some time trying to talk about how they are going to help manage the complexity of these 4-5 forks with a simpler app store distribution/composition frameworks and tools. That is yet another topic to cover in a fixed amount of time.
In short, there is even
MORE software to cover in a relatively fixed amount of time. That should tend to shrink the time on complete end system products covered. Unless , there is some huge synergy with the new software or Apple happened to finish at this time there is going to be even less time space on WWDC keynote for Apple end product demos.
