I’ve seen mention of M1 Max Duo and Quattro. That’s not a bad sounding solution.Kind of wonder though if Jade2C and Jade4C got cancelled in favor of just skipping the mostly M1 foundation and going on to M2 ( probably TSMC N4 based) one.
Minimally it is likely a different die layout than what the M1 Max uses. That "Max" is the other issue. How do they go "up" from the name 'Max'. What is bigger than the Maximum?
Could be M1 Max2 and M1 Max4 . Or perhaps get on a better name track with the M2 prefix.
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 Mac Pro Apple Silicon is going to eat our lunch
- Thread starter rondocap
- Start date
- Sort by reaction score
In a compact mATX-like box, in addition to MPX-like SoC daughter board and PSU, the rest can be all for PCIe expansion slots. Minimum five single-wide PCIe Gen 4 16-lane slots with max length ~260mm are no problem.
The usage for PCIe slots will be similar to current & past Mac Pro: audio interface cards, video capture cards & accelerators, solid-state disk expansion cards, super-fast networking cards. E.g. Apple could re-label some of the rejected M1 Pro/Max dies and create Afterburner Gen 2 cards.
I'm not sure Apple wants to kill off macOS x86-64 completely. Say in the very distant future (10+ years), Apple wants to/have to transition desktops back to x86-64...In the short term, there are various reasons, Intel Mac Pro will be available for a few more years. It's for [allowing] third-party [more time] to bring their expansion cards to ARM platforms, for a small number of users who need terabytes of DRAM, for users who need the latest & most powerful GPUs
I very much doubt that Classic Intel Mac Pro phenomenon will repeat itself. The phenomenon was formed on very specific conditions. Some of those simply will no longer exist in Apple silicon Mac Pro. New Mac Pro won't be affordable to mainstream to begin with. Parts won't be available in abundance and cheaply in second hand markets because no more cheap CPU, GPU, DIMMs thanks to PC world will be usable.
256GB embedded & soldered RAM is plentiful for lots of users. Next iteration will bring 512GB. Beyond that who knows what next DDR standard brings. I'm not sure Apple will add DIMMs to Mac Pro. They will likely not to if they can get away with it by keeping most users happy.
EDIT: in square brackets
Last edited:
If trying to fill up a 1700mm2 package faceplate with more letters perhaps makes more sense.
Max Duo and Max Quartet would be more consistent if going for entertainment grouping naming theme.
Latin's Quattuor would be even longer.
More think about it the more this won't necessarily be exactly a "Max" that is being doubled and tripled up. Apple could either dump the second Video de/encoder. ( would still have two and four in the combinations if each unit only had one. ). that would replace the "bottom" of Max chip with interchip comm. Or it would just be a bigger/different die.
M1 Double
M1 Quadruple
would be lots of letters too.
The "pro" and "max" are iPhone marketing left overs. M1P2 and M1P4 would work fine if willing to get off of reusing iPhone stuff. if this gets constrained to just one Mac product ( Mac Pro) then very few are going to buy it because of the name anyway.
Pro/Max dies have about three x1 PCI-e v1 lanes on them. Or maybe with some hackery x4 PCI-e v3. They would be pretty bad at being a computational engine on an add-in card where the answers had to go in/out over the PCI-e bandwidth available off the chip.
I wouldn't hold breath on an Gen 2 AFterburner either. As the transistor budgets get bigger that is going into the Apple SoC.
Unless there is some huge radical change in the x86_64 implementation space ... that's probably not going to happen. Apple is leaving and not likely going to backtrack here. A W-3300 series Mac Pro wouldn't be a sign Apple was going to backrack. It would more so be that they were not willing to commit fully to old Mac Pro space replacement with M-series. Apple can take a slice of that space and just leave the rest behind.
The longer that there are no other GPU drivers on the macOS on M-series side the more Metal turns into a proprietary graphic code "moat' around the software. The profit margins and inertia will get too large to let a x86_64 reversal happen. ( there is also likely a huge assumption here that GPU drivers will evolve on the x86_64 side over that 10 interim gap also... that too is likely illusionary if the driver lockout continues on the M-series side).
A Intel Mac Pro is needed to "bring up" a ARM card ? Errr how? The underlying boot foundation is completely different from T2/UEFI.
A M-series MBP with a thunderbolt PCI-e enclosure box would be a better match to do ARM boot environment card development with. A M-series Mac Pro with 1-2 slots even better.
Use latest , most powerful GPUs aren't going to do much of anything useful if Apple doesn't sign the GPU drivers. (which they are not for a wide variety of GPU cards. And that list is likley to get longer if they stay on the track they are on now. ).
The TB of RAM use case ? Yes.
Add-in card use case is more so for card that don't have any M-series drivers. ( probably will tip over to legacy cards and ones with minimal updates to stay current). [ and ones that work when native boot over into Windows (which is not likely for a M-series model ) ]
The most powerful GPUs are on a "bad" track for the Mac Pro 2019 .
ASUS confirms 12-pin PCIe Gen 5.0 connector for next-gen GPUs, announces new ROG Thor power supplies - VideoCardz.com
ASUS ROG Thor PSUs with PCIe Gen5 power connector ASUS confirms there is a new power connector for next-gen GPUs coming. The rumors started with our leak on GeForce RTX 3090 Ti have finally found a confirmation. It was alleged that NVIDIA’s next-gen flagship graphics card will be the first to...
The next gen double die , 400+ W monster cards are pretty antithetical to the direction that Apple is going. A W-3300 wouldn't necessarily help much because is approximately +50-80W step up in thermal drama too.
The "happiness" tension is likely to run in conflict with Apple RAM prices. The "Max4" (Jade4C) minimal RAM buy requirement would likely be 128GB. For someone who needs 32-40GB and buying 4x that amount at Apple prices... they will be lots of grumbling by many users.
The current, biggest use for expansion cards, is upgrading the graphics.
What you're describing is the 2013 Mac Pro with Thunderbolt 2.
Thunderbolt 2 did all the things (including with with PCI chassis) you're listing.
We've seen what the results are.
It's a failure.
There is no market for a "Pro" *workstation* with non-upgradable graphics.
Last edited:
We are all Apple Nostradamus, here is in my vision.
I think the next Mac Pro will be modular like the Intel's NUC Compute Unit, but MPX form factor, probably called Compute MPX.
The storage will be separate in a different Storage MPX.
When an upgrade comes out, one was just need to purchase a new Compute MPX to upgrade.
There would probably be supplemental Graphics MPX, to add additional GPU cores.
Also, Memory MPX for memory upgrades if not ready for a full Compute MPX upgrade.
I think the next Mac Pro will be modular like the Intel's NUC Compute Unit, but MPX form factor, probably called Compute MPX.
The storage will be separate in a different Storage MPX.
When an upgrade comes out, one was just need to purchase a new Compute MPX to upgrade.
There would probably be supplemental Graphics MPX, to add additional GPU cores.
Also, Memory MPX for memory upgrades if not ready for a full Compute MPX upgrade.
According to me, Apple will completely abandon the X86 and only propose Apple Silicon. In their eyes, xeons and AMD GPU are like G5 processors, too greedy and too little powerful. They will force the pros to convert to their codecs, their software and the software companies to adapt. Mac pros are already few-sold machines and maybe the mac pro 2019 has been designed just to show how the mac pro AS will be superior.
The most powerful GPUs are on a "bad" track for the Mac Pro 2019 .
ASUS confirms 12-pin PCIe Gen 5.0 connector for next-gen GPUs, announces new ROG Thor power supplies - VideoCardz.com
ASUS ROG Thor PSUs with PCIe Gen5 power connector ASUS confirms there is a new power connector for next-gen GPUs coming. The rumors started with our leak on GeForce RTX 3090 Ti have finally found a confirmation. It was alleged that NVIDIA’s next-gen flagship graphics card will be the first to...videocardz.com
The next gen double die , 400+ W monster cards are pretty antithetical to the direction that Apple is going. A W-3300 wouldn't necessarily help much because is approximately +50-80W step up in thermal drama too.
The PCIe Gen 5 power connector: 4 new pins for signalling. 4 additional power pins for a total of 12 pin delivering about 600W
If I recall correctly, MPX connector could deliver 500W. So now people have something to watch for when the refreshed 2019 Mac Pro comes out..
The differences are PCIe expansion slots will be housed inside the new Smaller Mac Pro. MPX-like SoC daughter board will be user replaceable.
SoC daughter boards will carry CPU, GPU, RAM in one package. Perhaps soldered SSD as system disk with additional SSD as socketed blades on Logic Board.
Reasonable to expect daughter boards from the same generation are interchangeable. Hence, upgradable. Remain to be seen if new daughter boards can be individually ordered from Apple.
Across successive generations, I'm not sure SoC daughter boards will be interchangeable. I think Apple has no reason to explicitly block this path. Neither will they guarantee cross-generation interchangeability. Sounds reasonable?
Yeah, my hunch is they won't release an update to Afterburner. But given they did an encode/decode ProRes engine in the new M1 Max, I'm hoping they can update the current Afterburner to do both encode/decode (however faint that hope is at the moment).
As I have a M1 Max MBP coming in already, I personally don't need that mentioned Mac mini. The allure of the afterburner is to continue to utilize my Curren 7,1 Mac Pro further, which is already upgraded to a degree and connected to all my other unlisted workstation components like external RAID arrays, multiple monitors, etc.
No, not at all.
It doesn't matter how good Apple Silicon is, in 12 months Intel and AMD are going to be, and stay, neck and neck with Apple. The Mac is just going to be "that computer you can't upgrade the graphics on". Nvidia are going to remain in their current position of utterly kerbstomping Apple's GPU options.
Microsoft will take the strategic decision to never release Windows for AS machines, even if they release an ARM version, its licence will explicitly block use on Apple hardware, native OR virtualised. The professional apps that came to the Mac because they could easily port their Qt/Intel codebases will go away, or demand Apple buy their continued presence the way Blender did.
In 5 years, Apple Silicon will be the new PowerPC.
Last edited:
That's a big fear for me too.
They might break too many things going for their own silicon and tightening their walled garden with allmost everything. They are consentrating too narrowly to only a few professional segments, audio and video. It might very well be things will go backwards in regards of software compatibility and availability, at least in regards of multi discipline professional use of Macs, as a cause of all this.
I am an architect and work in AEC field (Architecture, Engineeering and Construction). Now finally that we had the necessary support from necessary software companies at long last, mostly because of Apples intel transition I think, they might now have broke all that. That's what the PPC road means to me in this context.
SW companies most probably can't reuse their code as efficiently as they did with "only intel" era for Win and Mac, and Linux too in there in the same camp.
Ok I am not a programmer nor a software architect, so I don't really know for sure. Only thing I know is that some sw companies have not yet announced anything regarding ASi compatibility, their intensions or roadmaps. Most promising say they have started the work, but no timeline. Some say they are investigating it. Some just stay silent like a brickwall.
I can decide if I want to go through another transition still with Apple. It's just not that easy of a decision, and my needs professionally have changed through years, and lately too, by a lot.
But our Lunch though, that's gonna get eaten for sure.
And once Apple Silicon loses the edge, Apple will be fettered to a corpse, a dead weight that they have an institutional need to preserve, that they won't be able to ditch for reasons of pride - they'll be politically incapable of being processor agnostic and agile.
Not for nothing has the market bet that Microsoft has a better future for growth than Apple, and now values the company with a higher market cap.
Apple probably thinks that the Mac having software commonality with the iPhone / iPad is more important, than the Mac having commonality with the rest of the computing world. I suspect this will turn out to be misguided.
Last edited:
Yeah, this is also on my mind too. But I work with software that will almost always flow with these transitions (Adobe, etc.), so not as concerning per se.
I wouldn't be surprised if Apple continues to create builds of MacOS for Intel for some time to come, even after they sell nothing but Apple Silicon Macs. They have to for quite some time for now anyway. And this would allow for the opportunity to switch back, should it ever be warranted by Apple, as much as I don't see that happening, due to pride, etc.
MPX's 500W limit is more so given provisioning two , peak load , 250W GPU dies. Not a single 500W die. I would not crank up expectations that Apple is going to follow on the super TDP die path. It is possible; just not highly probable. ( Apple is exhibiting highly bigoted preference for their own GPU. And Perf/Watt is a laser focus. )
Is there any indication at all in the last decade that Apple is going to follow this path? ( Much of the 2009 design principles ... Apple isn't really following anymore) This reminds me of the 2017-2019 period where there as a rash of "Lego Mac Pro" suggestions. That Mac Pro would go hyper modular with CPU module , X , Y , and Z module that all snap together.
The SoC has the security processor in it. The SoC has both (in 2009 terms ) the "Northbridge" and "Southbridge" functionality to it. (e.g., all the USB port provisioning it out of the SoC ; not the "Southbridge". ) Every single I/O stream will be shuffled onto this "MPX-like" card. This is a much bigger entanglement that the 2009 system was (CPU + northbidge). All the video out. All the Thunderbolt out (and appropriate "retimers") . etc. etc.
Also going to run hotter than the 2009-2010 "CPU" cards did also. The RAM is soldered down so have daughtercard space constraints also. ( card can only be just so tall (vertical in the slot) and so long. )
[ For the 2009 boxes from the "Southbridge" and out the two systems Single socket and dual socket systems looked the same. That commonality drove the approach. There was lots of routing complexity that just stayed the same. If move almost everything to the card then not getting much 'reuse' there. It is more lots of redo and retime. ]
Reasonable? Doubtful.
i. for PCI-e v4 and up this will add to complexity and power. Extra socket traversals have overhead. Routing overhead. space that the 4 x16 lanes take up when coming out of W-3200 versus space that 4 x16 sockets soak up on the logic board.
ii. While the NAND daughter cards can e replaced in MP 2019 making it so that both sides of the SSD can be mixed/matched opens up even more issues. Even more hard resets.
iii. Likely gong to cause lost slot space if SoC heatsink is taller than it is wide. (i.e., 3D space utilization is lower )
iv. If the RAM is soldered down ( and Jade (M1 Max) die strongly suggests it will be) then will have "lost" the hyper modular customer base anyway.
Apple intel builds will likely be mostly targeted to the hardware that Apple sold in the past. Not random distant future GPUs off-the-shelf at microcenter/bestbuy. The primary objective to keep the systems not yet on Vintage/Obsolete status supported with some updates on "appropriate" set of features.
One of the major drivers of new GPU updates was other (non Mac Pro) systems getting new 3rd party GPUs over time. That is over. ( iMac family on down is being covered exclusively by Apple GPU ). External thunderbolt GPUs got covered as the new macs with updated GPUs got rolled out. Apple didn't chase off into the random GPU card market just for TB eGPUs and then do Macs as an afterthought.
Things like "hand off" that leverage Apple specific tweaks to Bluetooth/Wifi also tend to get localized to the newer hardware over time.
Over a longer stretch of time there highly likely will be stuff they don't "backport" to the Intel systems. ( there is stuff that T2 won't do. )
When you're talking about a board with the RAM/CPU/GPU/storage, what you're talking about isn't an exchangeable daughterboard, it's an exchangeable motherboard.
I don't know exactly what the cost/benefit is on having a AS Mac with a swappable main/everything board. Doesn't seem great to me. You're basically replacing everything but the case. Might as well just buy a new Mac Pro.
Storage at least could be factored out of the current soldered-board approach. So maybe it's replace-everything-except-for-storage. Still not a great deal.
Maybe plausible that the RAM is standalone and not soldered. But AS is so sensitive to the RAM type/number of channels/number of filled channels that it's hard to see that working across generations. There are even differences between Pro and Max chipsets that make them not compatible with each other's RAM configurations.
For what it's worth, I tested a maxed out M1 Max vs a maxed out 2019 Mac Pro with 28 cores, afterburner, and quad W6800x (two duos) and in Pro res and pro res raw, the M1 max was actually faster in almost everything - those new decoders and encoders are very capable.
Of course, with heavier GPU based codecs like r3d raw, the Mac Pro is still superior, but the M1 is catching up there too.
I predict the AS Mac Pro to be a monster, and beat the current maxed out Mac Pro for less cost.
Apart from GPUs and PCI-e SSD cards, are there other use case scenarios for expansions (thunderbolt 4 may not suffice scenarios ) ? If so, are these fields apple used to be/is/plans to be competitive in ?
Attended a tech conference a fews ago, actually, everything is transitioning to ARM. Many companies are developing their custom tailored ARM and moving away from general purpose x86 CPUs. ex. Amazon has been growing the use of their in-house ARM based processors running their AWS.
Mainly consumer, general use (ie. laptops, desktops), small businesses use general purpose x86 CPUs. All the top processing is going towards customer dedicated SOCs.
Sorry, but I don't understand all the talk of motherboard - soldered or not - RAM on conceptual Mac Pros. M architecture has shared memory with the GPU, using wide data paths.
It seemed to me when Apple introduced the Max processor, its difference to the Pro was data path and GPU transistors. The earlier M processor has 18 billion transistors; the Pro has 33.7 billion. The Max has 57 billion.
And these processors share the memory between them - and only up to 64 GB on the Max.
So IMO, the logical outcome for a desktop, is if you want to increase performance, you add in more of the Max processors. Who knows if Apple would build a Super Max which had 90 million transistors, it's increase being for GPU processing? They'd only be selling that to desktops, and most likely, as GPUs. It would not make much sense unless the same processor found its way into a notebook product.
It might be more sensible to design a motherboard with very wide data paths and allow each CPU to add the other one's shared ram. I think I've seen this approach with the Borg on Startrek ...
I reckon the retail price on the Max processor is $400 each. That is cheaper than many workstation GPUs, isn't it?
So if you bought a Mac Pro, I guess the question is, how many CPU's might it support? If it supported 8, then that would cost a customer 8 x 400 = $3,200 (for 8 Max processors). Who would also get 8 x 32 GB RAM = a quarter of a terabyte of RAM.
If you up the RAM, then Apple charges $400 for the extra 32 GB, per processor which relatively expensive.
For their current Mac Pro, Apple charges for two Radeon Pro (the proper pro vector suitable reliable etc) GPUs, with 32 GB each, erh hhh $11,600. The 9600 does have 60 billion transistors. Slightly more than a single Max processor. Compared to buying 8 Max processors, at my $400 each. For $3,200.
Apple charges $3,000 for 384 GB RAM on the Mac Pro. OK that's expensive compared to buying it yourself. But if one bought 8 Max processors, you'd get 256 GB of RAM included with them, for $3,200. But to get an extra 32 GB RAM with each processor, Apple charge $400 extra per processor on the Macbook Pros, so it would cost you $3,200 to get half a terrabyte of RAM, or double the 256 GB RAM. So you pay a fare bit for the extra RAM. Big deal I suspect if this model came into existence. But:
Surely the way to increase GPU performance on an M Pro Mac is via expanding CPU/GPU processors? They'd be a cheap way to do it if the motherboard could handle multiple processors.
It seemed to me when Apple introduced the Max processor, its difference to the Pro was data path and GPU transistors. The earlier M processor has 18 billion transistors; the Pro has 33.7 billion. The Max has 57 billion.
And these processors share the memory between them - and only up to 64 GB on the Max.
So IMO, the logical outcome for a desktop, is if you want to increase performance, you add in more of the Max processors. Who knows if Apple would build a Super Max which had 90 million transistors, it's increase being for GPU processing? They'd only be selling that to desktops, and most likely, as GPUs. It would not make much sense unless the same processor found its way into a notebook product.
It might be more sensible to design a motherboard with very wide data paths and allow each CPU to add the other one's shared ram. I think I've seen this approach with the Borg on Startrek ...
I reckon the retail price on the Max processor is $400 each. That is cheaper than many workstation GPUs, isn't it?
So if you bought a Mac Pro, I guess the question is, how many CPU's might it support? If it supported 8, then that would cost a customer 8 x 400 = $3,200 (for 8 Max processors). Who would also get 8 x 32 GB RAM = a quarter of a terabyte of RAM.
If you up the RAM, then Apple charges $400 for the extra 32 GB, per processor which relatively expensive.
For their current Mac Pro, Apple charges for two Radeon Pro (the proper pro vector suitable reliable etc) GPUs, with 32 GB each, erh hhh $11,600. The 9600 does have 60 billion transistors. Slightly more than a single Max processor. Compared to buying 8 Max processors, at my $400 each. For $3,200.
Apple charges $3,000 for 384 GB RAM on the Mac Pro. OK that's expensive compared to buying it yourself. But if one bought 8 Max processors, you'd get 256 GB of RAM included with them, for $3,200. But to get an extra 32 GB RAM with each processor, Apple charge $400 extra per processor on the Macbook Pros, so it would cost you $3,200 to get half a terrabyte of RAM, or double the 256 GB RAM. So you pay a fare bit for the extra RAM. Big deal I suspect if this model came into existence. But:
Surely the way to increase GPU performance on an M Pro Mac is via expanding CPU/GPU processors? They'd be a cheap way to do it if the motherboard could handle multiple processors.
Last edited:
A true SMP configuration probably would not work. The current series of designs all depend on everything being closely linked, and everything sharing the same address space and busses. There's other side effects too. The amount of RAM and the amount of RAM chips changes the amount of GPUs you can have. So you can't just slot more RAM chips in. You'd have to upgrade memory. Which also means upgrading the memory bus on the CPU. Etc etc.
Thats exactly why the Pro/Max configurations are what they are. Everything is so tied together that everything has to be upgraded together. To get the upgraded GPU, you have to get the upgraded CPU that's capable of addressing the upgraded RAM required for the extra bandwidth for the upgraded GPU.
It's really hard to see Apple building a system where you get to swap things out component by component. I would be like building a car where all the wheels are different sizes. There'd be no way to guarantee proper performance.
Yup, and PowerPC is still used on Mars rovers and other deep space vehicles, doesn't mean its relevant for desktop processing, which is what Apple, and everyone who uses computers in the same physical place as themselves, does.
Even if the entire rest of the industry, including Intel goes to ARM-derived designs, it's Apple Silicon that's going to end up as the new PowerPC, not ARM in general.
Yes. It seems to me the way to leverage economy of scale of your own production of the combined CPU/GPU/memory modules ie M, M Pro or M Max, is to multiply them on a motherboard designed to efficiently combine those resources. While the motherboard would cost a bit more money, the high value add components would be those mass produced - those modules. A side benefit would be the economy of scale for software development - developing for the M processors, would allow covering the whole gamut of Apple hardware. From even iPad perhaps to the highest end Mac Pro.
Curiously too, there is a lot of flexibility surrounding M Pro and max modules. For the same module interface, you can fit a 16 GB RAM Pro with 8 or 10 cores. GPU cores can be 14 or 16 33.7 billion transistors), or with the Max, up to 57 billion transistor with 32 GPU cores (also 24 available) . Ram per module 16 or 32 GB or on the Max 32 or 64GB. There's a lot of permutations & combinations with the choices available with the Pro and Pro Max processors (hopefully modules).
If Apple builds such a motherboard.
Last edited:
Sony/AMD created a "unified memory architecture" in PS4 ten years ago but AMD (or Intel) didn't attempt to bring the architecture to PC. Assume people in there knew it didn't make business sense. I don't know if it makes more sense today in PC world. We're likely to see first hints from PC companies soon.
Apple's current approach has no problem in laptops and AIO desktops. While I'm not as hyped as Apple Silicon sub-forum, I'm happy to see Apple brings "new" system-level architecture to PC computing. As far as we commons could see, Mac Pro is a problem for Apple's approach in which it doesn't fit and scale nicely. Mac Pro is also a very niche market segment for Apple. So normally won't receive tonnes of resources specifically targeting it.
I don't believe Apple is going to compete in general-purpose workstations. They're not aiming at the fastest CPU, most powerful GPU, and largest memory capacity on workstations. I remain cautiously optimistic and look forward to Apple bringing surprises. I believe a 10-year run with 3 to 4 generations of Mac Pro is no problem. They can transition desktops back to x86-64 after that if they want to or have to. Apple has no huge problem with Intel and there is also AMD - a risen star.
Apple indeed did it before or at least very similar in Class Intel Mac Pro where they had the CPU tray for housing the CPU and DIMMs. The new MPX-like SoC module will house CPU, GPU and RAM chips. These three items are inseparable in "Apple Silicon" approach. The new MPX-like connector will be mainly PCIe, power delivery plus some additional proprietary stuff.
The benefits are: 1) I imagine the SoC cooler will be huge. What could be a better form factor than tried and proven approach in Giant GPUs on the PC side? It makes efficient use of space and hence a compact mATX-like box overall. Not to mention it looks as cool as a MPX module. 2) other fringe benefits include easier manufacturing/SKU management/better serviceability & upgradability when compared to a single motherboard
Given new MBPs back to Titanium Book like-alike, 24-inch iMac (and new MBA rumoured) back in candy colors, isn't Apple in some sort of nostalgic? Of course, I'm no genius or fool at Apple. Your guess may be as good as many others.