Apple Silicon iMac & MacBook Pro expected in 2021, 32-core Mac Pro in 2022

Posted:
in General Discussion edited December 2020
Apple's short-term plan for Apple Silicon is reportedly a larger MacBook Pro and iMac refresh in 2021, with a Mac Pro with up to 32 high-performance cores coming at some point in 2022.

Apple's current Mac Pro
Apple's current Mac Pro


Backing up recent claims that Apple is designing an Apple Silicon-powered Mac Pro that is half the size of the Intel one, new reports say the company plans dramatically improved Mac processors. These include a potential 32-core Mac Pro in 2022, and an iMac in late 2021.

According to Bloomberg, Apple is expecting to produce Apple Silicon processors that are faster than even the current best performing Intel PCs. Citing unnamed sources said to be familiar with the plans, Bloomberg says that Apple is developing a series of processors with a series of options.

The sources say that Apple may choose to hold the fastest of its processors until 2022, and that it may release versions with varying numbers of cores. However, at present the plan for 2021 is a new MacBook Pro, and both an entry-level and high-end iMac.

The highest-end Mac Pro would be some time in 2022, and it may also benefit from greatly increased graphics capabilities. According to Bloomberg, the highest-end Macs could gain 64- and 128-core graphics processing.

Currently the M1 in the MacBook Air and 13-inch MacBook Pro contain either 7 or 8 cores. Future higher-end MacBooks, and what Bloomberg calls mid-range desktops, Apple is testing 16- and 32-core graphics processing.

Apple is rumored to have a large slate of products set to launch in the future, ranging from "AirTags" and "AirPods Studio" to an updated iPad mini. New Macs are also potentially on the way for early 2021.

These rumors include a 24-inch iMac, which could be the next Apple Silicon release following the initial three M1-based launches. Aside from the processor, the iMac is reckoned to have an edge-to-edge display, possibly borrowing design elements from the Pro Display XDR, and with an outside chance of a switch in display technology to use Mini LED.

Apple's next potential launch period for a Mac could be March 2021, with historical product launches tending to err away from January and February in favor of the third month. A December launch could still potentially happen, with one rumored for Tuesday, but it seems unlikely to take place so close to the end of the year.

However, given the last 12 months, Apple's decision to hold three product launch events in the fall, and its aggressive two-year timeline for the Apple Silicon transition, it's entirely plausible for new Macs to surface sooner rather than later.
«134

Comments

  • Reply 1 of 69
    That is insane!
    watto_cobra
  • Reply 2 of 69
    I presume this article is talking about 32 cores being on the same die. If true, that's good to hear. But don't see why Apple couldn't put multiple physical packages with N-cores each on the same system board. Moreover, Apple should do this as an expansion card for existing Intel Mac Pros so that Intel Mac Pros can run M1-apps natively.
    h4y3swilliamlondonviclauyyckillroywatto_cobra
  • Reply 3 of 69
    The iMac will come much sooner. It makes no sense for Apple to wait another year and release them end of 2021. 
    More likely, it’ll be released before Q3.
    razorpitwilliamlondonaderutterSamsonikkkillroywatto_cobra
  • Reply 4 of 69
    razorpitrazorpit Posts: 1,796member
    The iMac will come much sooner. It makes no sense for Apple to wait another year and release them end of 2021. 
    More likely, it’ll be released before Q3.
    Wondering if they are working on Intel emulators. There's a lot of us out here that still need to run the 64-bit version of Windows 10 Pro for specific reasons, and do not want to lose that ability. Maybe these 'higher-end' machines will get that function first? 
    killroywatto_cobra
  • Reply 5 of 69
    sflocalsflocal Posts: 6,093member
    razorpit said:
    The iMac will come much sooner. It makes no sense for Apple to wait another year and release them end of 2021. 
    More likely, it’ll be released before Q3.
    Wondering if they are working on Intel emulators. There's a lot of us out here that still need to run the 64-bit version of Windows 10 Pro for specific reasons, and do not want to lose that ability. Maybe these 'higher-end' machines will get that function first? 
    I bought a new iMac a couple months ago for this reason.  I require x(86) compatibility for my job.  No way around it.  By the time Apple  ends support for my 10-Core i9, 128GB, 8TB SSD system, Apple will have had a few years to fully-bake their ASI machines and hopefully, the issue of x86 compatibility will have been resolved, or I just worry about it then.

    I’m actually quite excited about these new machines.  Apple really blew past everyone else.  We may be seeing a new period of CPU designs, and seeing Intel becoming irrelevant.

    I do see these iMacs and to some degree, the Mac Pro losing much of the expandability options.  The iMac will be a giant iPad monitor and nothing more.  No way to open it.
    MisterKitpulseimagesentropyskillroywatto_cobra
  • Reply 6 of 69
    wood1208wood1208 Posts: 2,913member
    It will be relentless Apple Silicon assault on computer industry and no turning back. New performance bar will be set by Apple and Apple will break it. That story will continue in future.
    williamlondonaderutterMisterKitMacProSamsonikkkillroyradarthekatwatto_cobra
  • Reply 7 of 69
    rob53rob53 Posts: 3,251member
    Why stop at 32 cores. The current TOP500 supercomputer is comprised of a ton of Fujitsu A64X 48 compute core CPUs based on ARM v8.2-A. Cray is also working on changing to the ARM architecture. The current implementation is only running at 2.2 GHz using a 7nm CMOS FinFET design. This is on one chip! (ref: https://en.wikipedia.org/wiki/Fujitsu_A64FX, also https://www.fujitsu.com/downloads/SUPER/a64fx/a64fx_datasheet.pdf)

    Apple is already fabricating at 5nm and there's no reason why they couldn't build something similar, although I expect it to be a larger SoC than the current M1. The Fujitsu CPU appears to be more or less a standard size package, although it's only the CPU. (ref: https://www.anandtech.com/show/15885/hpc-systems-special-offer-two-a64fx-nodes-in-a-2u-for-40k)

    Apple isn't the only major computer company going ARM. It's good to see Apple finally expand to the Mac line. Those performance charts everyone's used to seeing with a simply curve are going to be amazed at how steep the jumps are once Apple really gets going. I ordered an M1 MBA simply because I've ordered ver 1 of several Apple Macs before. Crazy thing is this entry level Mac is faster than my current iMac. Talk about a steep performance curve!

    Note: I keep confusing ARM with AMD so previously I might have commented about the Fugaku supercomputer being AMD-based. If I did, my mistake. 
    williamlondonrazorpitviclauyycradarthekatwatto_cobra
  • Reply 8 of 69
    blastdoorblastdoor Posts: 3,282member
    32-core Mac Pro in 2022

    Saving my pennies for this one!
    radarthekatwatto_cobra
  • Reply 9 of 69
    mjtomlinmjtomlin Posts: 2,673member
    rob53 said:
    Why stop at 32 cores. The current TOP500 supercomputer is comprised of a ton of Fujitsu A64X 48 compute core CPUs based on ARM v8.2-A. Cray is also working on changing to the ARM architecture. The current implementation is only running at 2.2 GHz using a 7nm CMOS FinFET design. This is on one chip! (ref: https://en.wikipedia.org/wiki/Fujitsu_A64FX, also https://www.fujitsu.com/downloads/SUPER/a64fx/a64fx_datasheet.pdf)

    Those are designed for massive parallelism. This is not something that would be necessary for what is essentially a single user computer. You could get more perceived performance from a lower core, higher clock speed design.

    Even a 12 HP core design clocked higher would basically blow past almost everything on the market right now.
    dewmeradarthekatwatto_cobra
  • Reply 10 of 69
    Hopefully it will be cheaper than the Xeon Phi's were. 
    williamlondonwatto_cobra
  • Reply 11 of 69
    blastdoorblastdoor Posts: 3,282member
    I presume this article is talking about 32 cores being on the same die. If true, that's good to hear. But don't see why Apple couldn't put multiple physical packages with N-cores each on the same system board. Moreover, Apple should do this as an expansion card for existing Intel Mac Pros so that Intel Mac Pros can run M1-apps natively.
    Perhaps, but I suspect a chiplet strategy (see AMD Ryzen, esp Threadripper) is much more likely. (this used to be called MCM for :"multi chip model, see here https://en.wikipedia.org/wiki/Multi-chip_module

    As Apple has noted, the MBP 13, MBA, and Mini have historically represented 90% of Mac sales. That means the MBP 16", all models of iMac, and the Mac Pro make up the rest. I'll hazard a guess that the Mac Pro makes up far less than 1%. With such a tiny volume, I'd be surprised if Apple designed a single die just for that product. 

    With chiplets, you can design a small number of dies and then mix / match in a wide range of combinations to serve different markets. The combining doesn't happen with add-in cards to a motherboard (how quaintly 1990s of you!), instead they are integrated at the SOC package level. Just google AMD threadripper and look for pictures of the multiple dies integrated together. 

    Given the timeframe, Apple will likely use more advanced packaging than what AMD is currently using with Ryzen. I could imagine them designing three or four chiplets -- a CPU chiplet, a GPU chiplet, a central control chiplet (with things like the Secure Enclave, memory controller, SSD controller, etc.), and perhaps also an ML chiplet (or they could decide to integrate ML with one of the other chiplets). So a MBP16 and midrange iMac get control + 1 CPU + 1 GPU. High-end iMac could get 1 control + 2 CPU + 2 (or more) GPU. And then the Mac Pro could get even more CPU and GPU chiplets. I could also imagine, in high-end configs, using HBM3 RAM, at least as a monster cache. 

    I doubt there's a hard limit on how many chiplets could be combined -- it just means expanding the size of the SOC package. 
    edited December 2020 zimmiewilliamlondondewmewatto_cobra
  • Reply 12 of 69
    The timeline is roughly what everyone should have been expecting since the announcement. Low level Macs this year, higher end Macs next year and Pro Macs 2022. 

    Apple are sure to balance the need to bring to market as soon as possible with ensuring what they bring to market is adequate to replace the outgoing models. 

    Hardly anyone would,buy an iMac right now, so Apple need to bring out an iMac pretty soon, I just hope for a 27”/30” iMac and 16” MBP in Macrh 2021.
    bageljoeywilliamlondonwatto_cobra
  • Reply 13 of 69
    razorpitrazorpit Posts: 1,796member
    mjtomlin said:
    rob53 said:
    Why stop at 32 cores. The current TOP500 supercomputer is comprised of a ton of Fujitsu A64X 48 compute core CPUs based on ARM v8.2-A. Cray is also working on changing to the ARM architecture. The current implementation is only running at 2.2 GHz using a 7nm CMOS FinFET design. This is on one chip! (ref: https://en.wikipedia.org/wiki/Fujitsu_A64FX, also https://www.fujitsu.com/downloads/SUPER/a64fx/a64fx_datasheet.pdf)

    Those are designed for massive parallelism. This is not something that would be necessary for what is essentially a single user computer. You could get more perceived performance from a lower core, higher clock speed design.

    Even a 12 HP core design clocked higher would basically blow past almost everything on the market right now.
    I would agree with the current state of MacOS, but who's to say MacOS '11.1' wouldn't start pushing background processes to individual cores.

    Instead of applications sharing processor time on a smaller number of cores, you could theoretically scale it up for theses new beast-mode™ processors. I'm guessing the average user wouldn't notice much of a difference, but once you have it figured out it would be easier to apply it to the entire product range and have everyone benefit from it.

    I'm sure someone will come along and happily point out how wrong I am, but I'm just floating an idea out there. Remember there was a time when a large part of the "tech elites" also thought 64-bit processors on an iPhone was ludicrous.
    watto_cobra
  • Reply 14 of 69
    I presume this article is talking about 32 cores being on the same die. If true, that's good to hear. But don't see why Apple couldn't put multiple physical packages with N-cores each on the same system board. Moreover, Apple should do this as an expansion card for existing Intel Mac Pros so that Intel Mac Pros can run M1-apps natively.
    Why stop at Intel Macs? heck if they could put M1 on a PCI expansion card, Let all Windows PC's run Mac/iOS software and the world is your oyster!
    watto_cobra
  • Reply 15 of 69
    zimmiezimmie Posts: 651member
    razorpit said:
    mjtomlin said:
    rob53 said:
    Why stop at 32 cores. The current TOP500 supercomputer is comprised of a ton of Fujitsu A64X 48 compute core CPUs based on ARM v8.2-A. Cray is also working on changing to the ARM architecture. The current implementation is only running at 2.2 GHz using a 7nm CMOS FinFET design. This is on one chip! (ref: https://en.wikipedia.org/wiki/Fujitsu_A64FX, also https://www.fujitsu.com/downloads/SUPER/a64fx/a64fx_datasheet.pdf)

    Those are designed for massive parallelism. This is not something that would be necessary for what is essentially a single user computer. You could get more perceived performance from a lower core, higher clock speed design.

    Even a 12 HP core design clocked higher would basically blow past almost everything on the market right now.
    I would agree with the current state of MacOS, but who's to say MacOS '11.1' wouldn't start pushing background processes to individual cores.

    Instead of applications sharing processor time on a smaller number of cores, you could theoretically scale it up for theses new beast-mode™ processors. I'm guessing the average user wouldn't notice much of a difference, but once you have it figured out it would be easier to apply it to the entire product range and have everyone benefit from it.

    I'm sure someone will come along and happily point out how wrong I am, but I'm just floating an idea out there. Remember there was a time when a large part of the "tech elites" also thought 64-bit processors on an iPhone was ludicrous.
    The XNU scheduler already does this. POSIX processes and threads are kind of a headache for programmers to use, so Apple added Grand Central Dispatch in macOS 10.6 in 2009. Most of Apple's new APIs since then have been aggressively asynchronous. For example, URLSession/NSURLSession makes it harder to process returned data in the thread where you made the call than it is to process it in a completion handler block which gets scheduled on another queue. GCD also has a concept of priority levels, which tell it what kind of core to prefer for a given queue's tasks. User-interactive queues will usually send tasks to a high-performance core, while background queues will usually send them to a low-power core.

    Even so, single-core performance matters more to most users' perception of speed than parallelism. I expect the Mac Pro to have fairly extreme parallelism, of course, but it is also almost certainly going to have faster individual cores as well.

    I agree with Blastdoor, chiplets are probably the way it's going to go. GPU performance scales almost linearly with core count, and core count scales linearly with die area, but reject rate scales with the square of the die area. The main thing keeping Apple's GPUs from competing with AMD's or Nvidia's is die size, but big dies give you fewer parts per wafer, and fewer of those parts are functional. GPU chiplets would let Apple get to competitive core counts with small dies.

    The only concern after that is heat dissipation. Chiplets don't help spread the heat into more heatsink volume like a CPU and a separate GPU card does. That said, heat pipes are pretty good at getting heat from point A to point B, so it may not be too big a deal. As long as the total CPU+GPU power consumption is under 300W, the Mac Pro's existing CPU heatsink is plenty. With the performance and heat we've seen from the M1, I bet the Mac Pro's combined TDP will be maybe 180W. Probably less.
    razorpitwatto_cobra
  • Reply 16 of 69
    Rayz2016Rayz2016 Posts: 6,957member
    I presume this article is talking about 32 cores being on the same die. If true, that's good to hear. But don't see why Apple couldn't put multiple physical packages with N-cores each on the same system board. Moreover, Apple should do this as an expansion card for existing Intel Mac Pros so that Intel Mac Pros can run M1-apps natively.
    Why stop at Intel Macs? heck if they could put M1 on a PCI expansion card, Let all Windows PC's run Mac/iOS software and the world is your oyster!
    The only way this would work is if Apple charged the same price for the card as they did for the whole machine.

    There is no way Apple is going to burn its own house down like this, and then have to support every single PC in existence.
    seanjwilliamlondontmayauxiowatto_cobraCheeseFreeze
  • Reply 17 of 69
    ph382ph382 Posts: 43member
    rob53 said:
    Why stop at 32 cores. 

    I don't see even 32 happening for anything but the Mac Pro. I saw forum comments recently that games don't and can't use more than six cores. How much RAM (and heat) would you need to feed 32 cores?

    williamlondonwatto_cobra
  • Reply 18 of 69
    22july201322july2013 Posts: 3,571member
    Rayz2016 said:
    I presume this article is talking about 32 cores being on the same die. If true, that's good to hear. But don't see why Apple couldn't put multiple physical packages with N-cores each on the same system board. Moreover, Apple should do this as an expansion card for existing Intel Mac Pros so that Intel Mac Pros can run M1-apps natively.
    Why stop at Intel Macs? heck if they could put M1 on a PCI expansion card, Let all Windows PC's run Mac/iOS software and the world is your oyster!
    The only way this would work is if Apple charged the same price for the card as they did for the whole machine.

    There is no way Apple is going to burn its own house down like this, and then have to support every single PC in existence.
    Apple could sell the M1 chip as a separate part for "Windows for ARM" PCs, but continue to prohibit macOS from running on any other vendor's PCs. That requires less "support." But still some support. I think that would be a more palatable and profitable idea.
    williamlondonwatto_cobra
  • Reply 19 of 69
    blastdoorblastdoor Posts: 3,282member
    ph382 said:
    rob53 said:
    Why stop at 32 cores. 

    I don't see even 32 happening for anything but the Mac Pro. I saw forum comments recently that games don't and can't use more than six cores. How much RAM (and heat) would you need to feed 32 cores?

    https://www.anandtech.com/show/15044/the-amd-ryzen-threadripper-3960x-and-3970x-review-24-and-32-cores-on-7nm

    The 32 core Threadripper 3970x has a TDP of 280 watts on a 7nm process. It has four DDR4 3200 RAM channels.

    Based on comparisons of the M1 to mobile Ryzen, I would expect an ASi 32 core SOC to have a TDP much lower than 280 watts. 

    I bet a 32 core ASi SOC on a 5nm process could fit within the thermal envelope of an iMac Pro. 
    commentzillawatto_cobra
  • Reply 20 of 69
    tipootipoo Posts: 1,142member
    Just to add something, GPU core counts are all counted differently and meaningless across architectures. An Apple GPU core is 128 ALUs, say an Intel one is 8. 

    Seeing what they did with the 8C M1, the prospect of a 128 core Apple GPU is amazingly tantalizing, that's 16,384 unified shaders.
    Samsonikktechconcwatto_cobra
Sign In or Register to comment.