You're correct regarding the inversion of the thermal issues from that era to now. At one time the G4 and G5 were the fastest available, but Apple marketing still claimed that when it was no longer true. I wouldn't expect it otherwise.
Perhaps not in all cases. Macrumors reported on Bloomberg's story about Apple's autonomous car development. It seems the CPU will run hot:
...The chip's capabilities mean it will run hot and likely require the development of a sophisticated cooling system.
So they are using those instead of pianos.
Looks legit.
Unless you're developing it for a A64FX supercomputer (I'm interested in seeing what Cray can do what it). There are advantages to a common OS and platform from desktop to HPC.Well, A64FX might be ok for running it (although I would be surprised if M1 Pro outperforms it, even in SIMD code), but it would still make a terrible workstation. Building software would take ages.
Well, you are not supposed to use it in new software. But it works fine for legacy stuff.
Well, they could use it as a cabin heater since, I'm assuming, this will be an EV....Perhaps not in all cases. Macrumors reported on Bloomberg's story about Apple's autonomous car development. It seems the CPU will run hot:
The Apple car chip is the most advanced component that Apple has developed internally and is made up primarily of neural processors that can handle the artificial intelligence needed for autonomous driving. The chip's capabilities mean it will run hot and likely require the development of a sophisticated cooling system.
I'm honestly not sure what your point is here. Apple is going to support whatever Mac Pro is out for longer because Intel supports the Xeons that Apple uses for much longer than they typically do the Core iX processors (where X is either 3, 5, 7, or 9). Apple only drops Intel Macs when they can't do something fundamental to macOS's operation or when they feel they can't guarantee a smooth experience. Given this, I can't imagine that they won't start building more functionality that requires either Apple Silicon or an Intel Mac with a T2 chip. I think that will be the next major limitation in system requirements. But that's neither here nor there as far as the 2019 Mac Pro is concerned.My 2012 iMac has security updates ending in 2022.
2009 MBP I researched on had security updates until 2018.
From 2013 I'd want the Mac Pro to be refreshed every 4-5 years
So that would be
When the Mac Pro with Apple Silicon debuts keep selling the Intel Mac Pro until demand dwindles to <20% of Apple silicon.
- 2013 Intel
- 2017/2018 Intel
- 2021/2023 Apple silicon
2013 iMac Pro came out because customers demanded for a pro desktop with 2017 Xeon chips. Same reason why 2019 Mac Pro was shipped later as users wanted a tower Mac.
It personally bothers me that the refresh was done this way.
I'm honestly not sure what your point is here. Apple is going to support whatever Mac Pro is out for longer because Intel supports the Xeons that Apple uses for much longer than they typically do the Core iX processors (where X is either 3, 5, 7, or 9).
Apple only drops Intel Macs when they can't do something fundamental to macOS's operation or when they feel they can't guarantee a smooth experience.
Given this, I can't imagine that they won't start building more functionality that requires either Apple Silicon or an Intel Mac with a T2 chip. I think that will be the next major limitation in system requirements. But that's neither here nor there as far as the 2019 Mac Pro is concerned.
Apple will continue to sell the Intel Mac Pro for as long as it believes that people need them. The Mac Pro is not a huge seller, but it's still important to Apple's Mac business.
There is a rumored Mac Pro with Intel still in the works coming out alongside an Apple Silicon Mac Pro. Given the rate of development of Apple Silicon native apps, I don't think that this isn't an unreasonable thing for Apple to do. The Mac Pro is not the iMac or the MacBook Pro. It's a whole different animal and it's designed to work with workflows above all else.
My 2012 iMac has security updates ending in 2022.
2009 MBP I researched on had security updates until 2018.
From 2013 I'd want the Mac Pro to be refreshed every 4-5 years
When the Mac Pro with Apple Silicon debuts keep selling the Intel Mac Pro until demand dwindles to <20% of Apple silicon.
2013 iMac Pro came out because customers demanded for a pro desktop with 2017 Xeon chips. Same reason why 2019 Mac Pro was shipped later as users wanted a tower Mac.
In my (very) humble opinion, I reckon a lot of people are overthinking what Apple's strategy could be. My take is this:
M1 Pro/Max have a significant reduction in power usage, meaning that they can designed into a greater number of devices. This inherently benefits the customer when it comes to choice, especially as a great many high-end computing tasks can now be brought down to a more compact chassis.
The MacBook Pros have already demonstrated that there is virtually no compromise between two different sizes of device, therefore allowing the customer to think differently about their purchase. And even the M1, a consumer focused SoC, is more than capable for most users.
I think it's therefore safe to assume that Apple can differentiate it's desktop machines by function rather than performance, since it will become increasingly difficult to convince people to buy into a larger, more expensive machine.
Mac mini [M1, M1 Pro/Max] - The current SoCs would all perform within the existing chassis. I don't believe Apple would rename the higher-tier model anything else, for the simple reason that professionals and hobbyists are already aware of the mini's strengths: diverse I/O and size. It may end up being, along with the Mac Pro, the only Mac to have USB-A.
iMac [M1] - Consumer features that scale well within the chassis, designed mainly for a wireless user with a focus on consumption.
iMac Pro [M1 Pro/Max] - All speculation, but again we can safely assume a larger mini LED, enhanced speakers and more I/O will be present. Like the current 27" this is a very important product for Apple, as it satisfies users who need need internal expansion or modularity. I don't expect any 'duo' SoCs, for cost reasons and to push people to buy the Mac Pro.
Mac Pro [M1 Max Duo/Quadro] - Honestly, I envisage this just being a smaller 7,1 Mac Pro. Once you remove the MPX functionality, the double-height slots (single only) and reduce the power supply capacity, you've potentially taken away a third of the volume. PCIE support will still exist for the users who Apple tried to claw back with the current machine.
I don't see a 'cube' Mac and feel that messages may have been lost in translation.
2017 iMac Pro not 2013.Apple doesn't backport all the secuity updates anymore. Can handwave that this is "supported' because get something more than zero effort, but there is a really slacking effort being put into v n-2 operating systems at the end. The security updates are not timely nor necessarily complete.
when the 2013 Mac Pro came Apple stopped selling the 2012 ( really 2010 with a '2012' sticker slapped on it). If the Apple Silicon 'Mac Pro" debuts it will either be a replacement for the MP 2019 or a non-replacement that is given an overlapping name to proclaim transition completeness.
You're somewhat time continuity challenged to think the MP 2013 has anything like circa 2017 Xeon processors in it. It had more like tweaked 2012 ( E5 v2 versus E5 'v1' ) processors. The iMac Pro had 2017 processors as a lineage of a large fraction of what the MP 2013 market that system addressed. A large fraction of the folks who never got off the 2008-2012 models never considered the MP 2013 a viable option ( and that was largely detached from the Xeon version number ).
In the first case (basically complete overlap in functionality coverage), the MP 2019 would be retired and put on the support countdown clock. Just like 2012's disappeared.
However, The rumors point to a "half size" Mac Pro. It probably isn't a "replacement" of same basic MP 2019 capabilities. Some have said that "half' really means chopping on multiple dimension. With that "new math" definition of 'half', then it would even less be a replacement (probably have thrown general PCI-e add-in cards out the window).
The reason that the Intel would stick around wouldn't be on some percentage unit sales countdown clock. it woud more so because the replacement didn't really do the same 'job'. Even 50% of 1% is still a small number that would be in danger of being shutdown by Apple as 'too small to put effort into" product.
There is little evidence that Apple is trying to build an " Eypc, Xeon SP , Ampere Altra , etc. server baseline design"-killer SoC. If they are not then it will likely be difficult for them to do a direct replacement for the MP 2019. If they defcous from large scale general I/O then system won't cover the same functionality ground. There could be a decent double digit number of cores and a heft embedded GPU core count, but overall system would likely have very different characteristics on the I/O and RAM capacity support dimensions.
If Apple shipped another Mac Pro '2021-2022' that they could ride as a 'hobby product' for 2-3 years then the fab processes improvements they can fold into a 2024-25 SoC that still stock with largely embedded I/O would be in the "high enough" zone of coverage that they could just walk away in a similar fashion that the MP 2013 was a "walk away" from a substantive fraction of the user base. If supports 1-2 slots for non GPU cards then even more likely of having "just enough" to do a "walk away".
Mac mini, cube, cylinder, iMac (Pro) are all intended for external expansion OR be used connected to a network. There are many who wants that setup and it has nothing to do with devices being portable.I mostly think along your lines but I feel a dual m1 max maybe offered in the iMac Pro. If a closed Mac Pro system shows up, some peeps maybe tempted to go with the iMac Pro(neither would have expansion in any case)
Apple’s current ‘house’ design permeates a lot of their lineup and Apple tends to keep the design consistent for a few years and we are fairly early in the current design era..but Apple recently filed for a patent for an all glass cover design..a cylindrical trash can like device as well as something that looks like a mini Mac Pro was part of the patent diagrams (along with an iPhone). It maybe some years before it will see light of the day.
That may never materialize or come in some other material that looks like glass. The glass feel reminds me of the Mac systems of the early 2000s that had the sheathed in transparent glass vibes (the material was plastic though )
But that’s besides the point. What gives me the shudders is the return of the cylinder. Apple ate humble pie on the Mac Pro front and released an insanely pricey tower with “by definition modular” features to ..what ? Buy time ? ( the by definition modular comment was followed up with ‘that’s why we are releasing a pro display’..a clever choice of follow up words)
‘Thermal corner’ was just one of the issues that Apple faced with the tcMP so I wonder if it still makes sense to go down that route ? The ‘portable’ powerful desktop experiment failed miserably twice but in the two previous adventures they were reliant on 3rd party cpu and GPUs…Apple may be third time lucky this time around…or perhaps that’s what they feel… they may release a new pro display along with the cube/cylinder/whatever and call it - by definition - modular.
They could release it along with a refreshed intel tower and again - by definition - offer a choice. The apple market blitz could well spin this as the third coming of the Mac desktop.
The MPX module may well fit additional SOCs daughter cards and I do remember a big thread on the fate of the Mac Pro from a few years ago… someone had mentioned a clamshell design and ‘cartridges’…when the new Mac Pro arrived I assumed the cartridges’ are the MPX modules (no idea what happened to the clam shell rumour though)Mac mini, cube, cylinder, iMac (Pro) are all intended for external expansion OR be used connected to a network. There are many who wants that setup and it has nothing to do with devices being portable.
The cylinder "thermal corner" was 500W (same as iMac Pro) and Apple misjudged AMD and Intels abilities to produce low power draw chips. Therefore Apple should have never removed tower model from the lineup.
500W would fit 4 X M1 Max. If it makes commercial sense to have both a tower and cylinder/mac Mini Pro is up to debate.
I would vote for a 27 inch iMac with same internals as MBP and a headless Mac Mini Pro with 2X M1 Max and external displays for the pro user. The tower should be left as is with MPX modules accepting additional M1 chips as a cluster for compute. That is how the graphics cards are used today. cannot imagine why they did a MPX module just for the graphics cards. An MPX module draw 400W which equals the power draw of a rumoured 4X M1 Max assuming the power draw will be four time of a single M1 Max. What a coincidence...
Portability was touted as one of the desires of users who used the Mac Pros for on field work. Thars why you have those expensive wheels to move it around if needs be…There are many who wants that setup and it has nothing to do with devices being portable.
MPX modules are basically PCIe cards with additional PINs for power delivery if I'm not wrong, with has a max bandwidth of 32 GB/s. It is an order of magnitude less bandwidth than what the M1 Max can achieve today.The tower should be left as is with MPX modules accepting additional M1 chips as a cluster for compute.
Cube/xMac would be perfect for me. A single or double PCIe slot that could support a single proper exchangeable GPU, and a M1 Max, and I would be very happy. Of course, way below 5000$...I don't see a 'cube' Mac and feel that messages may have been lost in translation.
Cube/xMac would be perfect for me. A single or double PCIe slot that could support a single proper exchangeable GPU, and a M1 Max, and I would be very happy. Of course, way below 5000$...
Heck, a M1 Max Mac mini with eGPU support and I would be extremely happy.
But in general I agree, a Cube/xMac is not going to happen, the customers are too few.
It‘s time to buy a MacBook Pro Plus with M2 Quadro chipIn my (very) humble opinion, I reckon a lot of people are overthinking what Apple's strategy could be. My take is this:
M1 Pro/Max have a significant reduction in power usage, meaning that they can designed into a greater number of devices. This inherently benefits the customer when it comes to choice, especially as a great many high-end computing tasks can now be brought down to a more compact chassis.
The MacBook Pros have already demonstrated that there is virtually no compromise between two different sizes of device, therefore allowing the customer to think differently about their purchase. And even the M1, a consumer focused SoC, is more than capable for most users.
I think it's therefore safe to assume that Apple can differentiate it's desktop machines by function rather than performance, since it will become increasingly difficult to convince people to buy into a larger, more expensive machine.
Mac mini [M1, M1 Pro/Max] - The current SoCs would all perform within the existing chassis. I don't believe Apple would rename the higher-tier model anything else, for the simple reason that professionals and hobbyists are already aware of the mini's strengths: diverse I/O and size. It may end up being, along with the Mac Pro, the only Mac to have USB-A.
iMac [M1] - Consumer features that scale well within the chassis, designed mainly for a wireless user with a focus on consumption.
iMac Pro [M1 Pro/Max] - All speculation, but again we can safely assume a larger mini LED, enhanced speakers and more I/O will be present. Like the current 27" this is a very important product for Apple, as it satisfies users who need need internal expansion or modularity. I don't expect any 'duo' SoCs, for cost reasons and to push people to buy the Mac Pro.
Mac Pro [M1 Max Duo/Quadro] - Honestly, I envisage this just being a smaller 7,1 Mac Pro. Once you remove the MPX functionality, the double-height slots (single only) and reduce the power supply capacity, you've potentially taken away a third of the volume. PCIE support will still exist for the users who Apple tried to claw back with the current machine.
I don't see a 'cube' Mac and feel that messages may have been lost in translation.
No clue how they will connect these. Economy of scale will likely benefit a common SoC even if some parts (GPU/CPU/ML/Endoders) will not be used by all. As long as it is cheap enough compared to a dedicated GPU, no one will complain (ah well).The MPX module may well fit additional SOCs daughter cards and I do remember a big thread on the fate of the Mac Pro from a few years ago… someone had mentioned a clamshell design and ‘cartridges’…when the new Mac Pro arrived I assumed the cartridges’ are the MPX modules (no idea what happened to the clam shell rumour though)
Now Apple could well release daughter SOC mpx modules (and yes the current ones support 500w cards. No coincidence there. It was obvious with the modular talk and the subsequent release of said modules that’s how Apple saw expansions going forward.)
But if they do, we still don’t know how the system will see these modules :
Nodes ? Does that mean increased license costs ?
What interface ? PCI-e gen 5/6/7 etc or proprietary?
Besides will the end user be ok with extra CPU (or GPU) cores they may not need for their work ? Certainly it would be cheaper to sell the Mac Pro SOC cards as is with all bells and whistles intact ( no need to design extra options ). Apple could well price them good enough so it doesn’t matter whether your expansion needs CPUs or GPUs ( you get the other one free )
Compute via GPU is equally restrained regarding bandwidth so I do not see the issue. Send the job to the daughter boards which sends back a result. Of course this only works for highly parallel work where each solution is independent from each other.MPX modules are basically PCIe cards with additional PINs for power delivery if I'm not wrong, with has a max bandwidth of 32 GB/s. It is an order of magnitude less bandwidth than what the M1 Max can achieve today.
I would think Apple will have better results going wild with the mainboard design with their SoC.
The MPX modules will likely be reserved for storage or network expansions.
No clue how they will connect these. Economy of scale will likely benefit a common SoC even if some parts (GPU/CPU/ML/Endoders) will not be used by all. As long as it is cheap enough compared to a dedicated GPU, no one will complain (ah well).
Compute via GPU is equally restrained regarding bandwidth so I do not see the issue. Send the job to the daughter boards which sends back a result. Of course this only works for highly parallel work where each solution is independent from each other.
I would think Apple will get more bang for their bucks spending engineering resources making sure their AS Mac Pro motherboard has tons of high speed unified memory with tons of processing cores rather than trying to squeeze more performance out of a tiny PCIe channel for specialised work. Imagine 1-2TB of 800GB/s unified memory with a 40/128 CPU/GPU cores AS Mac Pro.Compute via GPU is equally restrained regarding bandwidth so I do not see the issue. Send the job to the daughter boards which sends back a result. Of course this only works for highly parallel work where each solution is independent from each other.
Who is certain? The MP is a challenge and therefor the most interesting one to follow. Does current MPX module support infinite fabric? Yet people put two GPU into it so the PCI bus must be sufficient for something.There is already a precedent here: paired GPUs with the Infinity Fabric Link. Metal has an API that tells you wich GPUs support fast data exchange so that you can write your apps taking advantage of this. A modular Mac Pro might work in a similar way.
I was more thinking of 120 CPU/384 GPU based on Jade 4C rumours.I would think Apple will get more bang for their bucks spending engineering resources making sure their AS Mac Pro motherboard has tons of high speed unified memory with tons of processing cores rather than trying to squeeze more performance out of a tiny PCIe channel for specialised work. Imagine 1-2TB of 800GB/s unified memory with a 40/128 CPU/GPU cores AS Mac Pro.
Who is certain? The MP is a challenge and therefor the most interesting one to follow. Does current MPX module support infinite fabric? Yet people put two GPU into it so the PCI bus must be sufficient for something.