The $1,099 4K iMac - if Apple does drop the price of entry to a 4K display, which makes sense at this point - would likely get a quad-core i5-8259U as found in the base 13" MacBook Pro, or my proposed $799 Mac mini. Then the current 1 TB Fusion drive which only has 32 GB of SSD space, and Radeon 550X dedicated GPU out of necessity.
So to get an iMac with comparable specs as my proposed $1,099 Mac mini (but a hex-core desktop
i5-8400 instead of a mobile
i7-8850H), you'd be looking at $200 more for the $1,299 tier, and possibly $100 more to get the 256 GB SSD.
$300 for a 4K display, Magic Keyboard, Magic Mouse, and a slightly better Radeon 555X GPU is not unreasonable. But it's also not what all Mac mini buyers would be looking for.
[doublepost=1533493134][/doublepost]
This is a company that used to care about consumers and products over shareholders. While that may not be true anymore, 4 GB soldered RAM in a 2018 Mac mini would be laughably bad. (4 GB upgradable RAM would be slightly better, but still not ideal for an out-of-box experience.)
[doublepost=1533493809][/doublepost]
The high-end model's logic board would be based off the 15" MacBook Pro, so perfectly doable from an engineering standpoint.
If Apple welcomes us back to 2012 standards, the answer is yes. The $999 Mac mini and $2,799 MacBook Pro with Retina Display shared the exact same CPU, the quad-core i7-3720QM.
Because of the dedicated GPU, I'd expect the 2018 version to be priced $100 higher, making it $1,099. Any higher would make it unreasonable when compared to the iMacs which include a built-in 4K display.
And the $2,799 MacBook Pro has a Touch Bar, twice as much RAM, twice as much SSD space, and a higher-end dedicated GPU in addition to the other things you mentioned.
Prior to that, the Late 2013 model iMac used a slightly slower desktop class i5 with integrated Iris Pro 5200 graphics (i5-4570R). That was a 65w quad core part like all of the other iMacs so subsequent base models have suffered a
downgrade in cores and performance.
i5-8259U is possible in a 2018 iMac but previous precedents have been set:
Base Late 2014 iMac 21.5": i5-4260U - 15w 1.4GHz - HD5000 graphics
Base Late 2015 iMac 21.5": i5-5250U - 15w 1.6GHz - HD6000 Graphics
Base Mid 2017 iMac 21.5": i5-7360U - 15w 2.3GHz - Iris Plus Graphics 640
All of these machines used a basic 1080p panel.
The late 2014 iMac uses the same CPU as the base 2014 Mini but lack of refreshes since have left the 2014 Mini behind while the iMac has relatively flourished with 2 updates, about to be 3.
The only question Apple need to answer is what constitutes an acceptable level of performance for a 4k panel if - as I expect - the base iMac goes 4k. Apple may decide that 4k won't get decent performance from a 15w CPU with UHD620. It might be that the Iris Plus 640 isn't good enough either and Intel don't have a successor product on the menu as far as I can tell anyway. The i5-8259U may have enough grunt but an i5-8250U + a dGPU certainly would be good enough even though there are precious few PCIe lanes on such a product to support one and have 2 good TB3 ports and PCIe SSD as well.
This is how you'd get your i5-8259U (28w) Iris Plus 655 Graphics into the base iMac but is it going to be at a cost that will allow Apple to make enough margin?
I'm afraid your calculations with respect to 'ideal' costs and specs are more than wishful thinking when put against real world Apple figures - what you find to be reasonable simply won't be offered by Apple.
It's not something I like to say but there's a long track record of Apple keeping to the same price points and making specs fit their profit margin - it's part of their marketing ethos.
Just go to the Apple website - plug the i5-8259U into the top SKU Mini at that price and with their Fusion drive plus TB3 ports and ask yourself if you like the sound of that after you've added 16Gb RAM and 512Gb SSD on the Mac ordering page. That's $1399 for that spec by the way.
I also think Apple couldn't build two motherboards to support an entry level CPU in that model because - once again - their engineering budget may not support 2 different motherboard designs. - it's been mentioned several times already in this thread. So we either get a Mini that's good at the top SKU or meets the price point at the low end SKU but probably not both.
This is the point which kills off your argument about the i7-8850U before we even get road to rationalising about your guess about how much Apple would charge for such a thing. I wouldn't expect much change from $2k if it happened but I think we're past the days when that might happen. Heck, here's my rebuttal anyway:
a. Apple won't release something that
will cannibalise the modular Mac Pro months before it's launched. Heck it would cannibalise the current 2013 Mac Pro once you spec it up. Imagine the Barefeats guys firing it up and declaring every other Mac short of the iMac Pro a goner in terms of bang per buck at '$1099'.
b. Phil Schiller has admitted that some professionals buy the Mini - I think he'd rather they bought a base model Modular Mac Pro if he knew it would come at a reasonable price. this is the marketing argument against allowing a high power mini before they launch the Modular Mac Pro. I've said the Mini - if viable - would change when the mMP does.
c. Putting another cooking CPU into the old 45w TDP case - with 6 cores - may result in more fried units if we examine the tests and benchmarks going into the 15" MBP at the moment. There's an engineering argument against a 45w CPU going back in that case - I guess you don't remember the complaints about the heat from quad core 2012s when they are being hammered by video rendering apps for hours. It just gets the rose tinted memories about upgradable RAM and drives.
d. UHD630 graphics would probably be seen as inadequate for driving a 4k display by Apple at the top end (note - inadequate - not incapable). Apple are long since past declaring the HD4000 ok since it went into the 13" Retina MacBook Pro but they might not mind for a base model.
e. Would Apple pay for re-engineering the Mini to fit the 6 core 45w TDP? Unlikely. That's the accounting argument. 4Gb in the base model Mini was laughable at time of release when we realised it would not be upgradable. It's a complete joke in the macOS Mojave era but yet the 2014 Mini soldiers on because it hits the magic $499 mark. Remember the 2012 Mini started at $599 back in the day so it must have been an accounting decision to do what they did.
Don't get me wrong, if the top end SKU looks attractive and comes with the i5-8259U if we do get a 2018 Mini I'd give it very serious consideration but I am in no way expecting the prices that you've bandied around on your threads. In fact price may be the only thing that puts me off it. The benchmarks on such a beast would be too good to be true but I fear that it means they don't think that product will cannibalise the modular Mac Pro because it'll be priced in the stratosphere.
There's still a possibility that we don't get a 2018 Mini because the professionals will get a base 2019 modular Mac Pro to feast on - if Apple can get the professional Mini users on a budget to swap over.
I'd still argue that favouring the upper end will increase average selling price of the Mini - making more profit per sale. Bean counters happy, marketing happy, professional/pro-sumer users happy. I'd argue that the average punter buying a $499 Mini isn't going to be buying much else from the Apple eco-system so it could only be there to cater to shareholders. The difference 4 years later is we have some very capable iPads now, capable of taking up the strain for $299-$399. There's room for a move upmarket.
We all know Apple won't be going back to 2012 standards - that would involve letting users update their own memory and HD.
The classic 2012 15" MacBook Pro had a dGPU along with the 15" screen and i7-3615QM and cost $1799 on launch (4Gb RAM etc). The equivalent Mac Mini cost $999. Not too bad in those days but modern Apple wants you to experience the convenience of a gorgeous screen without fiddling with your own low rent $99 non retina non widescreen monitor (overemphasised for clarity).
In engineering terms Apple may be able to make it, but the accounting and marketing guys will laugh them out of the first meeting if asked to put a $1099 ticket on it. In my opinion your spec is $1799 minimum and therefore not getting off the ground because nobody would buy the lowest SKU modular Mac pro if it started at $2499 as they would already have spent their budget.
A 2018 Mini released now would represent something that doesn't challenge the Modular Mac Pro - Apple not afraid of cannibalising mMP sales. And that for me says smaller, less power hungry (environmentally friendly), and cheap.
Penultimate note: a $999 top SKU Mac Mini includes Fusion Drive (1Tb HD + 128Gb SSD) - the 2014 is a pre-fusion-nerf product so 1Tb Fusion has 128Gb SSD, not 24Gb like in the iMac. The base $1099 iMac is still on a hard drive so it's harder to compare like for like even with the presence of a monitor. It becomes a very silly game later to compare later on this month if, as expected, the iMacs get Coffee Lake. The Mini would get obliterated in benchmarks if not updated at the same time.
Final note: I'm not sure how the accounting department at Apple will work - if they get a job lot of i5-8250U CPUs, they would put them in a new MBA successor, the base iMac (possibly with dGPU if it has a retina 4k screen) and they (again) have the opportunity to put them into a Mac Mini. As I mentioned a lot earlier they passed up 2 previous opportunities to update the Mini based on the base iMac. It'll soon be a third chance if the iMac goes all retina.
The only caveat I might have is how that fits with the rest of the range. Using the i5-8250U in the 21.5" iMac for me suggests the upper SKUs should be using something like the Kaby Lake G CPU to keep the theme of 4 cores, 8 threads CPUs consistent even though my feeling is that Apple wouldn't use those CPUs if they've already overlooked them for the MacBook Pros. That's my speculation for the day.
