I don't buy this report for the simple reason of the thermal issues they cornered themselves into last time.
A modular system can control for thermal better than an internally expandable one that could have no GPU or multiple beefy GPUs.
A mini only has to worry about the thermals for the CPU. The eGPU chassis has its own power and thermal design and limits.
But none of them can share thermal transmission or cooling components. They’d each have to employ some self-contained cooling. Maybe fine, if we aren’t talking about fans in every module. But adding more fans? Redundant component repetition between modules?
Apple hates fans and noise. I do too, but I accept the need for them to operate sometimes, and I’m not afraid of LARGE quiet fans. Apple is terrified of devices that aren’t stupidly thin.
Since Apple continues to make ridiculously thin and poor thermally-managed devices, this idea runs the risk of being multiple modules that each die from bad thermal design. Oh boy, we can replace a module without trashing the whole computer... talk about reinventing the wheel.
I’m hoping this is just another bullshit myth that gets dispelled, but I absolutely do NOT trust Apple. They clearly don’t care about reliability of the software (the tons of bugs I report every iOS release, none of which get fixed), and they clearly don’t care about longevity of the hardware under heavy use (nor the ability to run them at capacity for longer than a couple of minutes before being throttled).
Today’s Apple is not the Apple that gave us the fantastic products of the past. That expertise is gone and the corporate ideologies and priorities have changed. We really have no reason to trust this won’t be another head-in-the-sand, myopic, pile of ignorant arrogance.
It’s also too late for general feedback. Apple has done whatever Apple is going to do this round. If they screw up this time, it will just further compound the same long-term problem of which they already struggle to even be fractionally aware.
No straight forward tower with interchangeable off the shelf parts for you!
You mean just another expensive Apple product that does little more than a PC counterpart? What’s the point in that? I can almost hear the howling from the haters as Apple proves them right.
I don't buy this report for the simple reason of the thermal issues they cornered themselves into last time.
A modular system can control for thermal better than an internally expandable one that could have no GPU or multiple beefy GPUs.
A mini only has to worry about the thermals for the CPU. The eGPU chassis has its own power and thermal design and limits.
Wouldn't it make more sense to build in a thermal "chimney" so that a single fan module could control the airflow through the entire stack? That is, if someone were actually designing something like this. Otherwise I don't see how a tight stack of hot-running modules would make any sense. Each one would have to be blowing air horizontally through their own chassis.
H-P offers a somewhat modular compact desktop with different plug-in modules connecting through a USB connector. The Elite Slice has been on sale for a couple of years.
I do not know if the report is correct, but Apple has a long history of proprietary connectors used to lock customers in.
Some of Apple’s connectors provided benefits that Intel wasn’t offering at the time. If it wanted functionality not offered by it had no choice but to develop its own connectors. For example FireWire and the Lightening Connector. FireWire was way ahead of competing options. Some of its functionality has as far as I know still havent been Replaced. Apple was also among the first to adopt USB so I don’t think it adopts proprietary cables to lock people into anything..
I'll believe it when I see it. Has the distinct whiff of bullshit.
I have big, big concerns about bandwidth on this entire concept.
That's a legitimate concern and has always been so since the beginning of computer architecture and design, i.e., what's the interconnection cost versus computing cost in a multiprocessing architecture? For course-grained or isolated multiprocessing like you'd see in a data center or server farm the modular approach is inherently more scaleable for running a lot of jobs/loads in parallel, e.g., job-level MIMD. But if you want to apply a lot of processing power to a fewer number of jobs/loads you want to keep the communication cost as small as possible, e.g., job-level SIMD. Previous generations of computer (and microprocessor) architectures have adopted hybrid connectivity strategies in attempts to strike a balance between suitability for different types of workloads within a single architecture by having multiple communication busses/networks available, e.g., VME and VSB. Additionally, how the full-stack memory architecture is designed plays a huge role in how well a specific architecture fits multiprocessing workloads. A modular design with physically separate memory subsystems, e.g., a stack of Mac Minis, would have some distinct advantages over a multiprocessor architecture with a single shared memory (including caching) architecture depending on the workload.
If Apple is truly buying into a modular and extensible architecture I'm less concerned about raw bandwidth out of the gate than with Apple picking the right interconnect technology and architecture, at least at the granularity of whatever they consider a "computing node" or "computing element" to be. I'd prefer that they land on a standard technology like 100 Gigabit Ethernet ring network for synchronous and isochronous interconnections between nodes even if they use proprietary connectivity/bus technology for sub-elements within nodes. If they pick a technology that has an upgrade and growth pathway that is not exclusive to Apple they could have a much longer lasting design and also leverage partner capabilities. The Trash Can Mac Pro was a very innovative design but it was already at the end of its growth lifecycle the day it shipped. It had nowhere to go but down no matter how much Apple pushed it. I'd rather Apple release a design that has an upside and Pull from the market, other Apple partners, and customers.
It will be very interesting to see what imperatives/qualities are
driving Apple's modular architecture. Are they seeking better
scaleability in performance, better system serviceability and
upgradeability, better dynamic adaptability to differing workloads
(performance, task types, energy efficiency), more pricing flexibility,
better component reuse, fault tolerance and availability, or some
combination of many of these things?
This is all based on an old patent of Apple's, where they considered a stackable design. People want to be able to put in their own full size PCIe 3.0 video card, and memory and storage and this plan won't allow that. It learns little from Apple's failures with the current Mac Pro design.
Good grief, if I am unable to use industry standard graphics cards and RAM modules and instead they are incased in consumer hostile closed modules it will be time to just build an amazing Windows tower for my heavy lifting and a Mac Mini for all of my personal day to day stuff. I would hate to have to do that, and hope this is BS, but this is getting ridiculous. I really just want a current cheese grater.
Apple are one of the most stubborn companies around. Your cheese grater idea is very likely a better idea, so I’m sure Apple will enevitably disappoint you when they invent something cool, way overpriced and not practical for most pros. I’m sure it’ll look cool on Apple.com though 😉
I don't understand all the doubting Thomas's on this thread. Seems like a great idea.
As long as the connectors have enough bandwidth I don't see how it wouldn't work. Thermal management can be customized to each module, pulling in cool air from the sides and shooting it out the back. I'm willing to bet that multiple, small fans, running at partial strength, is much quieter than one or two giant fans pumping air through a giant rectangle; and in this case fans would only be used when needed, not the case with a large rectangle like the cheesegrater. Also, a thick rectangle doesn't make a lot of sense given that the hardware components are largely 2 dimentional; an array of two dimentional modules does make a lot of sense, however.
And of course the other obvious benefits.. highly customizable in function and cost, high performance capability, and it will lookamazing.
Good grief, if I am unable to use industry standard graphics cards and RAM modules and instead they are incased in consumer hostile closed modules it will be time to just build an amazing Windows tower for my heavy lifting and a Mac Mini for all of my personal day to day stuff. I would hate to have to do that, and hope this is BS, but this is getting ridiculous. I really just want a current cheese grater.
Apple are one of the most stubborn companies around. Your cheese grater idea is very likely a better idea, so I’m sure Apple will enevitably disappoint you when they invent something cool, way overpriced and not practical for most pros. I’m sure it’ll look cool on Apple.com though 😉
Of course, we should keep in mind that the cheese grater itself was considered "cool, way overpriced and not practical for most pros." the day it came out. They were always way more expensive than any of their PC counterparts. There is still a chnace Apple will do that again, but still end up making something great that people want to buy.
I'll believe it when I see it. Has the distinct whiff of bullshit.
I have big, big concerns about bandwidth on this entire concept.
If Apple is truly buying into a modular and extensible architecture I'm less concerned about raw bandwidth out of the gate than with Apple picking the right interconnect technology and architecture, at least at the granularity of whatever they consider a "computing node" or "computing element" to be. I'd prefer that they land on a standard technology like 100 Gigabit Ethernet ring network for synchronous and isochronous interconnections between nodes even if they use proprietary connectivity/bus technology for sub-elements within nodes. If they pick a technology that has an upgrade and growth pathway that is not exclusive to Apple they could have a much longer lasting design and also leverage partner capabilities. The Trash Can Mac Pro was a very innovative design but it was already at the end of its growth lifecycle the day it shipped. It had nowhere to go but down no matter how much Apple pushed it. I'd rather Apple release a design that has an upside and Pull from the market, other Apple partners, and customers.
It will be very interesting to see what imperatives/qualities are
driving Apple's modular architecture. Are they seeking better
scaleability in performance, better system serviceability and
upgradeability, better dynamic adaptability to differing workloads
(performance, task types, energy efficiency), more pricing flexibility,
better component reuse, fault tolerance and availability, or some
combination of many of these things?
All apple should care about is the software interface. They can market the interface with very little hardware. The Hardware is nice but all that matters is the screen, case and input devices (which are replaceable). The Interface brings the customers and in turn the developers. The interface drives all the current revenue the hardware is nice but the best products the hardware works really hard to just get out of the way.
Not to say that they should abandon hardware but if the new MacPro can be a sexy looking screen with all the noisy parts 50m away then the noisy parts can be anything and everything. They can be open hardware, they can be be whatever GPU suits. They can be whatever interconnect works best. They could also sell a case (not unlike the cheese grater) with said modules inside.
Well to me at least the Key the MacOS future is to get the processing happening away from the user in a seamless way. It's the last part Steve Jobs if I ran Apple agian speach that is still missing. The MacPro seems to be the best test for it. It is the customers with the demand for the noisiest hardware and the quietest enviroment.
I don't understand all the doubting Thomas's on this thread. Seems like a great idea.
As long as the connectors have enough bandwidth I don't see how it wouldn't work. Thermal management can be customized to each module, pulling in cool air from the sides and shooting it out the back. I'm willing to bet that multiple, small fans, running at partial strength, is much quieter than one or two giant fans pumping air through a giant rectangle; and in this case fans would only be used when needed, not the case with a large rectangle like the cheesegrater. Also, a thick rectangle doesn't make a lot of sense given that the hardware components are largely 2 dimentional; an array of two dimentional modules does make a lot of sense, however.
And of course the other obvious benefits.. highly customizable in function and cost, high performance capability, and it will lookamazing.
But it isn't a great idea - it fails on many points that Apple claims to take seriously:
It would be highly inefficient from a materials usage standpoint. Not green at all to make multiple chassis' with multiple fans, multiple power supplies etc. etc. etc. Lot's of unnecessary use of metals and plastics. Apple makes a big point about being green and manufacturing things efficiently.
Would it look amazing? A bunch of boxes stacked on each other is not sleek or elegant at all. If it used cables to connect modules, that is visual clutter and a mess. Even the cheese grater had a certain visual panache that a stack of modules will never have. Just doesn't fit the Apple design idiom.
Reliability, troubleshooting and support could be complicated by such a design. A multi-unit system would add a lot of variables that wouldn't exist in a single enclosure. Multi-modules and connectors present multiple failure points. If the connectors are embedded in
the external case and plug in as you stack, that also presents many
failure points and a place for dirt or spilled liquids to collect. As people have already said more than one type of connector would be needed for different bandwidth needs. How is that going to work? Wouldn't that dictate what modules could be stacked where? If it used cables, cables are notorious failure points in almost everything electronic. Apple is known for great support, but why would they build something that would make their job more complicated and expensive to implement?
It would without question be more expensive for users to build out a system, but unnecessarily so. You would be paying for a bunch of manufactured repetition that would not exist with a single enclosure. Apple doesn't like inefficiency - either in energy use or design.
Unless Apple has come up with a completely new spin on this sort of modular stack design, seems it would be a fail right out of the box.
Not buying this. Too obvious and not innovative in the least just like a Mac Tower.
I think the ‘modules’ will be smaller sticks with a low power footprint. The A12X package fits comfortably on an m.3 card. 10 billion transistor budget would make a decent GPU, a decent NPU or a decent dedicated CPU (if you exchanged the integrated graphics for extra cores) with room to spare for RAM/Flash on the stick.
I don't buy this report for the simple reason of the thermal issues they cornered themselves into last time.
A modular system can control for thermal better than an internally expandable one that could have no GPU or multiple beefy GPUs.
A mini only has to worry about the thermals for the CPU. The eGPU chassis has its own power and thermal design and limits.
Wouldn't it make more sense to build in a thermal "chimney" so that a single fan module could control the airflow through the entire stack? That is, if someone were actually designing something like this. Otherwise I don't see how a tight stack of hot-running modules would make any sense. Each one would have to be blowing air horizontally through their own chassis.
You mean like the trash can MacPro? I have one and it is awesome but...
The holy grail of unnecessary bespoke chassis and cooling nightmares.
The convenience of the stackable modules is wholly out of whack with how often you'll need to be modifying the stack, and would cause huge downstream problems that are way, way more detrimental than convenience.
Form over function to the nth degree.
Therefore the design is confirmed, no?
Obviously not, I already said I'll believe it when I see it. I have faith that Apple are smarter than this, the rumour is a flight of fancy from someone without a clue.
Comments
A mini only has to worry about the thermals for the CPU. The eGPU chassis has its own power and thermal design and limits.
Apple hates fans and noise. I do too, but I accept the need for them to operate sometimes, and I’m not afraid of LARGE quiet fans. Apple is terrified of devices that aren’t stupidly thin.
Since Apple continues to make ridiculously thin and poor thermally-managed devices, this idea runs the risk of being multiple modules that each die from bad thermal design. Oh boy, we can replace a module without trashing the whole computer... talk about reinventing the wheel.
I’m hoping this is just another bullshit myth that gets dispelled, but I absolutely do NOT trust Apple. They clearly don’t care about reliability of the software (the tons of bugs I report every iOS release, none of which get fixed), and they clearly don’t care about longevity of the hardware under heavy use (nor the ability to run them at capacity for longer than a couple of minutes before being throttled).
Today’s Apple is not the Apple that gave us the fantastic products of the past. That expertise is gone and the corporate ideologies and priorities have changed. We really have no reason to trust this won’t be another head-in-the-sand, myopic, pile of ignorant arrogance.
It’s also too late for general feedback. Apple has done whatever Apple is going to do this round. If they screw up this time, it will just further compound the same long-term problem of which they already struggle to even be fractionally aware.
Oh wait...
As long as the connectors have enough bandwidth I don't see how it wouldn't work. Thermal management can be customized to each module, pulling in cool air from the sides and shooting it out the back. I'm willing to bet that multiple, small fans, running at partial strength, is much quieter than one or two giant fans pumping air through a giant rectangle; and in this case fans would only be used when needed, not the case with a large rectangle like the cheesegrater. Also, a thick rectangle doesn't make a lot of sense given that the hardware components are largely 2 dimentional; an array of two dimentional modules does make a lot of sense, however.
And of course the other obvious benefits.. highly customizable in function and cost, high performance capability, and it will look amazing.
It would be highly inefficient from a materials usage standpoint. Not green at all to make multiple chassis' with multiple fans, multiple power supplies etc. etc. etc. Lot's of unnecessary use of metals and plastics. Apple makes a big point about being green and manufacturing things efficiently.
I think the ‘modules’ will be smaller sticks with a low power footprint. The A12X package fits comfortably on an m.3 card. 10 billion transistor budget would make a decent GPU, a decent NPU or a decent dedicated CPU (if you exchanged the integrated graphics for extra cores) with room to spare for RAM/Flash on the stick.