They'll only "begin appearing" this year though, and I would expect as is currently the case the Xeons would probably be later targets versus the consumer processors.
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
I was on 13.2 until a few minutes ago, when I updated to 13.3. The Lightroom lag is worse now. Damn.
Actually, the server CPU's would likely be the first to get updated silicon.
Even a ~3% speed downgrade in a server CPU can become costly.
Consumer CPU's have little chance of being compromised by any Spectre/Meltdown malware.
Last edited:
Yeah, after the latest info from Intel I’m pretty convinced that the Cascade Lake Xeon CPUs will have this fix. It would be weird to only implement the fix in the Xeon SP and not in the the Xeon W versions, so that is one more reason to look forward to the new Mac Pro.
Feisty, aren't we ?
Even if the numbers were correct, and assuming the higher priced iMac Pros will sell at all , Apple would have to sell a lot of upgrades to make it worth their while .
Especially since OSX needs to support every single GPU upgrade option properly, not something Apple is famous for .
Hence an open design - including some aspects of OSX - might or might not lose them a few bucks at the initial purchase, but gain them a lot more in units sold and market share .
Not to mention upcoming generations ; in the movies nerdy kids might do their fancy programming on some scruffy laptops .
Yet in real life, the first approach to serious computer work, and training, happens on the most affordable, most flexible and accessible hardware available .
[doublepost=1517059446][/doublepost]
I can't fully agree .
I'd like to point out the need for a headless Mac; an integrated display is very limiting in many applications - to put it mildly . The iMac Pro is extremely 'niche' because of that, at least in the more expensive configurations .
The classic iMac clientele appreciate the form factor and simplicity over everything else, yet still want some decent power - only most of those won't pay 'iMP' prices .
In that respect, the mMP will have to make the iMac Pro line virtually redundant for most of their targeted buyers .
Now that might be a hard pill for some people at Apple to swallow .
Damn shame you cannot attach one or more additional monitors to the iMac Pro, isn't it? Oh, wait.
And even if Apple does release a 32" 8K display in conjunction with the new Mac Pro, I expect the significant majority of purchasers will also be connecting another monitor as well - quite possibly a 27" 5K display like that on the iMac Pro.
If they are already paying iMac 5K prices for the Core i7, 32GB of RAM and a 1TB SSD because their workloads make use of those specifications, the jump to the base iMac Pro is not that ridiculous and the performance boosts of the faster CPU and GPU might improve those workloads even more.
This assumes the Mac Pro is just an iMac Pro without the display - so W-Series Xeons, 128GB of ECC and dual SSDs with a Vega 56 or 64 GPU on a card instead of soldered to the mainboard.
The real "risk" for a lot of folks on this board who just want to be able to swap video cards every 12 months is that this new Mac Pro is a "Tier One" workstation with the options and price-list to match. People are having heart attacks over the $5000 to $13000 price tag of the iMac Pro. Well an HP Z8 that is often put forward as to what the next Mac Pro should emulate starts at around $5000 for a configuration that the base iMac Pro will outperform and can be pushed to the near-side of $80,000.
Last edited:
Develop that, I actually dont undestad your point.
Disagree, Apple is used to "train" their customers to buy their products the way they sell, Apple's premium in propietary hardware pays itself, Guys from PC world used to DIY upgrades may see it inconvenient but Fanboys (and corporate) customers like this way, I come from PC (ok, first mac, the pc, and now Mac again) initially was my thinking, to purchase bare minimal configurations and then upgrade as I need, I do that on my first MBP(ram+storage+wifi/bt), and the MP(ram), Mini (storage), But also I'm aware DIY upgrade sometimes are risky(I burned logig boards upgrading GPUs), even time consuming (compatibility issues), nothing this happens if you just buy original sys->sell outdated original sys->buy new better spcd system.
A game changing Approach would be a Cartridge-like GPU solution, one with Apple's blessed compatibility and high quality std, plus simple KISS install, one that do not need you to spend time solving issues, or risk to system damage.
I read very carefully Apple's signals about the mMP, and points a Modular system (segregated display at least), upgradeable (At BTO or after market ?), and ALL NEW proudly Macintosh (which means innovative, groundbreaking, beautiful).
More strong rumor point Apple working with AMD on a Full AMD based MAC, maybe Apple delayed the mMP release to adopt the Epyc cpus (8/16/24/32 cores) and TB3 to be free from Intel (intel released TB3 to public domain on Jan 1), OK maybe the MP to be Xeon based, and this AMD cpu (an APU too) goes to the next Mac Mini (VR capable).
Personally I doubt Apple to resign the Mac Pro peripheral control (allow DIY PCIe upgrades), It means Support Nightmares (very expensive to apple), more work on macOS compatibility (as 3rd party GPU vendors will require support for its drivers), a long ETC that doesn't profits to Apple.
The GPU Cartdrige (or something alike) is an very strong possibility on the Next mMP, but I wont be surprised, if Apple just re-launches the Trashcan just grown enoutg to host dual 350W GPU plus a 32core unique CPU, Upgradable only at Build time, and maybe requiring authorized service just to upgrade RAM (as the iMac Pro), mathematically it will require an about 37% bigger trash can, or something alike.
I was very surprised how Apple didn't include a little window to allow DIY ram upgrades on the iMac Pro, and it actually is an not a good signal for the Mac Pro 7,1.
Last edited:
I am not. I think Apple will adopt the same internal cooling solution for its next iMac with 6 core, 64 GB max, Vega 56 with a couple of IOs less and perhaps the same screen as the iMac pro. It might be a little cheaper maxed out than the base iMac pro.
If the reports of iMac pro cooling performance are true, no reason why apple wouldn’t adopt it to the regular iMacs. There is the issue of the 21 inch variety though, so maybe not.
Those Xeon ws are utter opportunistic move from intel by creating some artificial segregation to sell lower quality chips while charging 2x for 25% better performance for the gold series. And I wouldn’t be surprised if Apple had a hand in that move with its thermal requirements for its iMac Pro.
A BTO Xeon w 18 core from HP adds $5530 when the suggested retail price is around $2500. A BTO Xeon 6154 gold 18 core ( at 25% faster base clock ) is $5990/- ( around the same suggested retail price by Intel )
Notice the BS ?
Last edited:
I believe the iMac Pro is a single-season product, at least with 18 cores xeon, maybe Apple later 'rethink' the iMac Pro as with an single AMD APU (the server APU with 8-16 cores and 24-32 GPU cores)., and name Pro a glorified iMac, the mMP will say a lot of the Mac future, specially its Architecture, moving to AMD cpus (a fact IMHO) means x86 architecture to last another 10 year, keeping Intel CPUs maybe a clear indication Apple later to switch to their own ARM cpus.
I doubt Apple is that keen to jettison the x86 platform ( it would love to if possible ) but not until it’s ipads reach Mac book Pro level performance. Or when you can develop for iOS using iOS’s hardware.
Most likely it would embed some arm cpus that can communicate with x86 Mac hardware, to allow running iOS apps on Mac hardware and it will be the low hanging fruits at first along with porting over some features from iOS devices - Face ID etc.. it is also easier to use a mouse/pencil for touch based based interface than the other way around. If Apple can get its iOS devices to run on Macs ( using arm chips ), it can extend the appeal of macs for many developers because they can develop once but target two ( and more if you add the watch and tv OSes ) platforms at the same time.
Apple has made some moves in that direction by unifying it’s file system for both platforms as well as adding a rudimentary file management system for iOS devices.
But it will likely be a ‘hosting iOS on Macs’ scenario rather than a full blown switch. That switch - if it occurs ( and likely will if Apple can find a way ) - is a few years away at least.
Coming back to iMac pro... I have no idea what the sales are, or if people are waiting for the iMac refresh before making a decision, or how happy people are using it, but if nothing else I think Apple has found the internal hardware structure for the iMacs of the future. If sales are good then I wouldn’t be surprised if Apple announces an 8k variety down the road to further differentiate between the standard ones and the iMac pros.
Reg AMD, I am, as I have stated earlier too.. open to a system based on the rather successful zen platform ( else no way would you have seen an 8 core CPU from Intel in mainstream markets )...
Provided two things happen : 16-24 core thread ripper single+multi core speeds ( no less ), option to add nvidia gpus that aren’t bottle necked by thunderbolt speeds, some commitments to timely release of updates .. zen+, zen2 etc
Hopefully developers will make use of AMD GPUs now that we have a reasonable card ( power inefficient though ) but maye Navi might solve that.
What else ? Shake the workstation market !!
Most likely it would embed some arm cpus that can communicate with x86 Mac hardware, to allow running iOS apps on Mac hardware and it will be the low hanging fruits at first along with porting over some features from iOS devices - Face ID etc.. it is also easier to use a mouse/pencil for touch based based interface than the other way around. If Apple can get its iOS devices to run on Macs ( using arm chips ), it can extend the appeal of macs for many developers because they can develop once but target two ( and more if you add the watch and tv OSes ) platforms at the same time.
Apple has made some moves in that direction by unifying it’s file system for both platforms as well as adding a rudimentary file management system for iOS devices.
But it will likely be a ‘hosting iOS on Macs’ scenario rather than a full blown switch. That switch - if it occurs ( and likely will if Apple can find a way ) - is a few years away at least.
Coming back to iMac pro... I have no idea what the sales are, or if people are waiting for the iMac refresh before making a decision, or how happy people are using it, but if nothing else I think Apple has found the internal hardware structure for the iMacs of the future. If sales are good then I wouldn’t be surprised if Apple announces an 8k variety down the road to further differentiate between the standard ones and the iMac pros.
Reg AMD, I am, as I have stated earlier too.. open to a system based on the rather successful zen platform ( else no way would you have seen an 8 core CPU from Intel in mainstream markets )...
Provided two things happen : 16-24 core thread ripper single+multi core speeds ( no less ), option to add nvidia gpus that aren’t bottle necked by thunderbolt speeds, some commitments to timely release of updates .. zen+, zen2 etc
Hopefully developers will make use of AMD GPUs now that we have a reasonable card ( power inefficient though ) but maye Navi might solve that.
What else ? Shake the workstation market !!
Last edited:
Removing the 3.5" HDD is what really allowed Apple's engineers to improve the cooling capacity of the iMac Pro by allowing the placement of the two fans and the center exhaust. So unless Apple drops Fusion Drives and goes to only SSDs, I can't see the iMac's cooling system being significantly changed.
That being said, the step-up from a 1TB Fusion Drive to a 256GB SSD is "only" $100, so Apple could just raise the base price by $100. Or perhaps they meet folks halfway and raise it by $50 since I expect most folks would then need to budget another $100 for a 4TB MyBook as 256GB probably won't be enough storage for most folks.
SSDs are the future for everyday use. If you are investing in an iMac , chances are you might be ok getting an external disk for long term storage.. esp if you also rely on time machine for regular, fire-and-forget backups. Investing in external archival storage is Always a wise move ..so even if you relied on fusion drives, hopefully you are backing up that data somewhere else. Besides what’s the point of all those thunderbolt ports anyway.
I think newer iMacs might come with 512gb as standard ... ssds have been coming down in price, esp for lower capacities.
Yes, Apple really should just drop the HDD / Fusion option from the 27" iMac line. Leave it on the 21.5" to meet a price point, but if you're ready to buy a 27", paying an extra $100/200 for a 256GB/512GB SSD as standard should not really be a make-or-break pain point.
I don't think we're getting rid of fusion drives until you can get a 512GB SSD as a standard; aside from penny-pinching I think Apple guesses (probably rightly) that low-end customers are going to care about running out of space a lot more than a slower system.
I think the iMac Pro's thermals coming to the regular iMac is a given, but I agree it'll have to wait for fusion drives to be phased out. I'm not hopeful on that front—my guess is this year will be the year hard-drive only models go away, but fusion drives will probably hang on for a while yet.
At this point they could also just use SATA instead of PCIe for the low-end SKUs to save money while still giving customers a faster system. A 500GB Samsung SSD is $150 these days, compared to the $500 cost for a NVMe M.2 module.
Both iOS and macOS now run on APFS (even first was on the iPhone) https://developer.apple.com/library...Management/Conceptual/APFS_Guide/FAQ/FAQ.html
(Emphasis added.)
I don't think that means what you think it means . . .
A "fact" is not reasonably subject to debate; no one knows if Apple will or will not move to AMD cpus, your "humble opinion" notwithstanding.
(Emphasis added.)
I don't think that means what you think it means . . .
A "fact" is not reasonably subject to debate; no one knows if Apple will or will not move to AMD cpus, your "humble opinion" notwithstanding.
Dear Good sir.
As it is 2018, standards in epistemology, logic, reasoning are no longer efficacious for usage in common dialogue. Might I recommend getting comfortable with yielding rhetoric, charisma and celebrity in your toolkit instead. Looking to the future, you may want to start practicing replacing all classical lingual lexicons with emojis, swipe gestures and photo filters.
We look forward to your compliance.
Thank you.
Well, the team I just inherited is one of these 'corporate umbrella' ones. My insight will follow below.
And the implications of this are that when my main guy comes to me with his exotic purchase list, my eyes glaze over and I go "YUP".
Well, in actuality, I'll ask him a few smart-sounding questions (such as what resolution of projector's he's buying for the next two octagons) and get his assurances that his purchase list will do what's supposed to be done and is kosher with our buying rules ... then I'll say "I'll check into it" .. then I actually don't, but wait 48 hours and then approve it anyway, because I know that I don't know this stuff. I expect it to be to the tune of $100K in the next 3-4 months.
Well, upper management wants to triple the size of this work business unit within the next three years, so its pretty safe to say that the next two years are going to be bigger than this year. In the meantime, I'm trying to sort out the corporate hiring system's rules & arcane rituals thereof, as well as figure out how to add the office space to accommodate them without losing a conference room...
While I know that that's a specific use case, where I see this going for the more freelancer (self-funded) customer is that the solutions need to not be gratuitously expensive. That's what I personally hit with the cMP to tcMP change-over for local storage: the prospects of paying the premium for a TB-based external stack of HDD's vs just tossing HDDs into internal bays resulted in a quite rude "Apple Tax" that grew the effective cost of a "Same Capability" system by ~25%.
While from a small independent that resulted in a "I'll get by on my cMP for awhile longer", in a corporate environment, that would have resulted in buying 4 Macs instead of 5 Macs ... not because Apple got more expensive (although they did), but because the money that was being spent on the third party vendors went up by 40%, and a new Mac without those externals is a futile purchase.
Simplified, the point here being that when Apple's system design decisions makes the workflow-needed system peripherals too expensive, Apple sees a decline in the sale of their base units (i.e., Apple loses sales).
eGPU performance hits aren't that drastic, not sure why it keeps getting brought up. I can see being upset about the principle of the thing (depending on your setup it's not as cost-effective or it takes up more room) but actual performance is a pretty minor consideration.
And no one seems to question why those decisions were made. Did the resulting new design function better than the old one ? Was it more cost effective ? Did it solve a long standing serious issue with the old design ? Why were things removed/locked down but ended up locking their own ability to update ( this one is a riot !) ? where did all those promised peripherals with thunderbolt disappear ?
Try finding a cost effective dual/quad eGPU solution. It costs almost as much as decent 4x GPUs themselves. Not to mention there is overhead of driver issues. I mean ...say I want to use 3 GPUs to render. I spend money to buy 3 TB3 enclosures. They end up costing same as 1.5 GPUs themselves. I have now got 3 power cords to hookup somewhere, 3 TB cables too. Also does the software know how to distribute the same render task to these three separate GPUs over TB?
I am not against eGPUs. But I think they are a solution looking for a problem... perhaps they can enable older systems to have acsess to VR..or play some games...but that’s where their utility ends. And it won’t be long.. it has almost started..where all new systems will meet the basic requirements of entry level VR..assuming it’s a pressing need at the consumer level. What will these eGPUs do then ?
Last edited:
eGPUs will still be faster than the entry level GPUs of those new systems. I haven't yet seen a case where a benchmark has run up against the actual limits of TB3 in these instances; you're losing some performance in the process but you're still going to get more out of a eGPU with a desktop GPU than you will a mobile solution. So laptop users who want more punch when they're docked and users with multiple computers they want to hot-swap will continue to find it useful. Obviously it's not the solution for some users, and if you're someone who is only working on a sitting-in-place workstation it's not really a feature that's for you.
Look at it this way. I can get tri/quad gpu setups in my workstation AND still get a mobile eGPU to hook up if needs be ( which again raises the issue of lugging those boxes + cords around with your mobile setup ).. until these eGPUs deliver power over a single cable ( ideally hooked into the laptop it self ) and small enough (25-30% bigger than the GPUs themselves ) I don’t see them as massively adopted for workflows... right now they seem to me as a stop gap arrangement... until mobile GPUs can handle basic requirements that these eGPUs can do at present that is .. or until there is some solution coming from a tangent I can’t see yet.
Last edited:
Because a Mac with a (no matter how small) compromise in performance due to requiring an eGPU for its serious GPU, will cost significantly more than a PC of similar performance with a slotted GPU, and when VR is the realm of consideration, performance hits both matter more, and macOS being "nicer" doesn't come into the picture, because you don't interact with the host OS. A Mac in VR is just a lower-quality experience for more money.
Apologies if someone has posted this before but i couldn't read 300 pages of posts.
Whilst procrastinating in the office I imagined a concept for the modular mac pro based on Apple's (Tim's) recent design philosophy:
1) make is small and thin
2) Solder everything to prevent users from upgrading the machine themselves
3) Add plenty of Apple Tax so Tim can make even more money
4) Use the smallest SSD/ RAM/ number of USB-C connectors possible in the entry level configuration whilst still calling it PRO
5) makes Lots of dongles! so that even after spending $5K you still need to buy more dongles and have a desk full of wire spaghetti before you can do any work.
Whilst procrastinating in the office I imagined a concept for the modular mac pro based on Apple's (Tim's) recent design philosophy:
1) make is small and thin
2) Solder everything to prevent users from upgrading the machine themselves
3) Add plenty of Apple Tax so Tim can make even more money
4) Use the smallest SSD/ RAM/ number of USB-C connectors possible in the entry level configuration whilst still calling it PRO
5) makes Lots of dongles! so that even after spending $5K you still need to buy more dongles and have a desk full of wire spaghetti before you can do any work.