Apple Silicon transition may hit its two-year target with 2022 Mac Pro

2»

Comments

  • Reply 21 of 37
    Gurman believes rumors about a final Intel Mac Pro are true, with one more apparently planned for launch. Rumors from July point to a Xeon W-3300 family CPU being used in an update, in parallel to Apple Silicon. 
    It makes sense to release a "final" Intel-based Mac Pro at the same time or slightly earlier than the Apple Silicon Mac Pro. There will be some highly CPU-optimized software out there (think vertical markets) that will not make the transition to Apple Silicon in a timely manner, because re-doing those optimizations is extremely expensive and time-consuming. Supporting these Intel-captive customers long enough for them to transition is a good idea.

    Keep in mind that there are people out there with decades of experience in optimizing code for x86 platforms and AMD/nVidia GPUs, but far fewer with that sort of experience on ARM in general and by defiition nobody with that much experience with Apple Silicon.
    edited August 2021 tenthousandthingswatto_cobra
  • Reply 22 of 37
    crowleycrowley Posts: 10,360member
    doggone said:
    mcdave said:
    The longer they’re leaving it, the more the competition has stepped up. The buying public can’t see beyond marketing specs so genuine advantages are already mitigated. Single-core performance has been matched by Intel 11th gen i7/i9 so it’ll be interesting to see how much Apple has left in the tank.
    Apple silicon is a paradigm shift for processors.  The speed advantages and the headroom for growth are massive and Intel won't be able to keep up.  This is just the beginning for Silicon Macs. 

    Intel will be able to survive because Windows is not going anywhere and non-Apple ARM chips are way behind Apple. Eventually they will get better and unless Intel can learn to develop ARM chips successful they will be left behind.
    Where's the massive headroom for growth?  

    The M1 is already at 5nm, while Ice Lake is at 10nm, so from that angle it looks like Intel has more scope for miniaturisation.
    Apple are already invested in big.little architecture, while Intel has only recently started exploring that area, so from that angle it look like Intel has more opportunities too.
    Apple's architecture is all in on RISC, while Intel are again only in the relative early days of moving parts of their instruction set to a static length instruction set for the associated benefits of hybrid RISC/CISC, they have avenues for development and improvement there.

    I'm not trying to poo-poo Apple here, I'm sure they can do great things, but not sure where this certainty about massive headroom has come from.  Intel processors are competitive with Apple's in many ways, superior in some, but with the notable exception of power consumption, and they have numerous avenues to improving that, if they can just whip their engineering into touch.  Plus, if Apple won't sell their silicon then they have a disadvantage in that their chip design will never be able to pay for itself, it will always be subject to Apple products, and their limited market presence.  Intel sell to everyone.  

    I wouldn't count Intel (or AMD, to whom 99% of the above also applies) out yet.
    muthuk_vanalingammobird
  • Reply 23 of 37
    zimmiezimmie Posts: 609member
    crowley said:
    michelb76 said:
    imagladry said:
    mike54 said:
    After all the unlimited praise youtubers, tech sites, Apple fanboius, etc gave the M1, I just hope Apple is not taking advantage of this praise, milking as much revenue as they can from it, thereby delaying advancement. Apple does have a bad habit of releasing something great and then sitting on it past its use-by date.
    ah, they produce a new A Series chip every year. My guess is that will be the plan for the M Series, also. Is that "sitting on it past its use-by date?"
    Probably, as the M1 is simply an A14X. 
    It's "simply" something that doesn't exist?  The definition of "simply" must have changed since I was a kid.
    If you look at the A12 versus A12X and A12Z, the changes are almost identical to the changes from the A14 to the M1. CPU goes from two fast cores to four fast cores, GPU goes from four cores to eight cores.

    If you look at the packaging, they're almost identical there, too. The big reason for the M1's limited memory performance and capacity is that it only has space for two RAM packages.

    Thus, the M1 can reasonably be called an A14X.

    crowley said:
    Intel processors are competitive with Apple's in many ways, superior in some, but with the notable exception of power consumption, and they have numerous avenues to improving that, if they can just whip their engineering into touch.
    Have you been following the Atom cores? They started as a minimal implementation of the amd64 instruction set without speed boosts like speculative execution and branch prediction. Once they had the basics working, they refused to accept changes unless each percentage point of additional power budget produced a greater number of percentage points of performance improvement. The recent cores are really capable. Power consumption is still a bit higher than ARM's, but dramatically better than other amd64 cores. They're also small enough to fit 50+ cores on a die, which is what the Xeon Phi line was.
    edited August 2021 watto_cobra
  • Reply 24 of 37
    zimmie said:
    crowley said:
    michelb76 said:
    imagladry said:
    mike54 said:
    After all the unlimited praise youtubers, tech sites, Apple fanboius, etc gave the M1, I just hope Apple is not taking advantage of this praise, milking as much revenue as they can from it, thereby delaying advancement. Apple does have a bad habit of releasing something great and then sitting on it past its use-by date.
    ah, they produce a new A Series chip every year. My guess is that will be the plan for the M Series, also. Is that "sitting on it past its use-by date?"
    Probably, as the M1 is simply an A14X. 
    It's "simply" something that doesn't exist?  The definition of "simply" must have changed since I was a kid.
    If you look at the A12 versus A12X and A12Z, the changes are almost identical to the changes from the A14 to the M1. CPU goes from two fast cores to four fast cores, GPU goes from four cores to eight cores.

    If you look at the packaging, they're almost identical there, too. The big reason for the M1's limited memory performance and capacity is that it only has space for two RAM packages.

    Thus, the M1 can reasonably be called an A14X.
    So the question is, then (if the rumors are correct and it is a thing), what will the M1X (or A14XX, if you will) be? Just seems like new territory here. Correct me if I’m wrong, but I don’t think there are any true parallels in the A-series history for this “XX” iteration, let alone a third “XXX” iteration as is being speculated on here for a Mac Pro.

    I’ll guess it’s likely they’ll follow the familiar A-series pattern with the M series, so there will be an M1X soon and then an M2 and M2X next year; but that leaves out the Mac Pro. So if there is going to be a third, highest-end, iteration, whether “Z” or something else, kind of seems like it might need to wait for the M2 architecture. 
    edited August 2021 watto_cobra
  • Reply 25 of 37
    Expected to be smaller in size, roughly half that of the current Mac Pro, the Apple Silicon version is anticipated to use chips with higher core counts, possibly including 20-core and 40-core variants.

    The next 18-months are going to be a truly painful moment for INTEL and AMD. Half the size? While size may not matter (ha) being able to shove a more powerful computer in half the space with better cooling and energy efficiency, is a very visual embarrassment for INTEL 
    and AMD. If Apple made a server version the reduction in rack space and cooling would be very attractive for server farms.
    williamlondonwatto_cobra
  • Reply 26 of 37
    flydogflydog Posts: 1,094member
    mcdave said:
    The longer they’re leaving it, the more the competition has stepped up. The buying public can’t see beyond marketing specs so genuine advantages are already mitigated. Single-core performance has been matched by Intel 11th gen i7/i9 so it’ll be interesting to see how much Apple has left in the tank.
    Wow, what an absolutely clueless post that won't age well. 


    watto_cobra
  • Reply 27 of 37
    robabarobaba Posts: 226member
    mbmoore said:
    Which states banned gaming computers? The same ones that mandated 100% electric cars by 2035? What morons. Power consumption only matters with laptops, tablets, and cell phones. Desktop power consumption is almost irrelevant in comparison as long as there is sufficient cooling.
    ::cough cough:: rollingblackouts ::cough::
    williamlondonMplsPwatto_cobra
  • Reply 28 of 37
    flydogflydog Posts: 1,094member
    Bloody hell… Hope this doesn’t mean we won’t see deliveries of M1x (M2?) MacBook Pros and MacBook Airs until 2022. Really counting on 2021 Q3 or worst case Q4 deliveries. 
    Why would you be counting on that?  Apple stated unequivocally that this would be two-year transition. 
    watto_cobra
  • Reply 29 of 37
    flydogflydog Posts: 1,094member
    Roderikus said:
    OMG 2 years - how hard is it to migrate iMac & MacBook Pro internals into a slightly larger casing ?!
    Yeah because that's ALL that project entails.  Is your theory that Apple is intentionally holding back new Macs in an attempt to make a smaller profit?
    williamlondonwatto_cobra
  • Reply 30 of 37
    doggone said:
    mcdave said:
    The longer they’re leaving it, the more the competition has stepped up. The buying public can’t see beyond marketing specs so genuine advantages are already mitigated. Single-core performance has been matched by Intel 11th gen i7/i9 so it’ll be interesting to see how much Apple has left in the tank.
    Apple silicon is a paradigm shift for processors.  The speed advantages and the headroom for growth are massive and Intel won't be able to keep up.  This is just the beginning for Silicon Macs. 

    Intel will be able to survive because Windows is not going anywhere and non-Apple ARM chips are way behind Apple. Eventually they will get better and unless Intel can learn to develop ARM chips successful they will be left behind.
    Really? A paradigm shift? It isn’t like it utilizes quantum optics to do it’s calculations.
    williamlondon
  • Reply 31 of 37
    flydogflydog Posts: 1,094member
    mbmoore said:
    Which states banned gaming computers? The same ones that mandated 100% electric cars by 2035? What morons. Power consumption only matters with laptops, tablets, and cell phones. Desktop power consumption is almost irrelevant in comparison as long as there is sufficient cooling.
    No state has "banned gaming computers."  That is utter bullshit.  California and some other states enacted regulations limiting idle power draw. Does not ban any computer, nor does it even limit power consumption when the machine is actually being used. 
    williamlondonAlex_Vwatto_cobra
  • Reply 32 of 37
    crowleycrowley Posts: 10,360member
    zimmie said:
    crowley said:
    michelb76 said:
    imagladry said:
    mike54 said:
    After all the unlimited praise youtubers, tech sites, Apple fanboius, etc gave the M1, I just hope Apple is not taking advantage of this praise, milking as much revenue as they can from it, thereby delaying advancement. Apple does have a bad habit of releasing something great and then sitting on it past its use-by date.
    ah, they produce a new A Series chip every year. My guess is that will be the plan for the M Series, also. Is that "sitting on it past its use-by date?"
    Probably, as the M1 is simply an A14X. 
    It's "simply" something that doesn't exist?  The definition of "simply" must have changed since I was a kid.
    If you look at the A12 versus A12X and A12Z, the changes are almost identical to the changes from the A14 to the M1. CPU goes from two fast cores to four fast cores, GPU goes from four cores to eight cores.

    If you look at the packaging, they're almost identical there, too. The big reason for the M1's limited memory performance and capacity is that it only has space for two RAM packages.

    Thus, the M1 can reasonably be called an A14X.
    It can more reasonably be called an M1, since that's its name.  But ok, I get your point.
  • Reply 33 of 37
    robabarobaba Posts: 226member
    Marvin said:
    Roderikus said:
    OMG 2 years - how hard is it to migrate iMac & MacBook Pro internals into a slightly larger casing ?!
    It depends on what they plan to do with the Mac Pro. It wouldn't make much sense to only put the exact same iMac/MBP chip options in the Mac Pro enclosure so they'd have to manufacture a new chip layout. That takes planning, design, testing, manufacturing, marketing, rollout, it's not a trivial process to make an entirely new chip.


    Not too sure about that last part.  If they use the same M1X SoC designs in a high performance SiP, the on-package power/data fabric should in fact be enough to “glue” the logic units into an effective cpu.  Easier said than done, but that’s what they appear to be attempting.

    but before that tackle this, they’ll want to get the M1X out the door and shook-down for any errata.  The M1X is in production now, so hopefully  the M1Z (?) will be on the bench soon.
    watto_cobra
  • Reply 34 of 37
    mattinozmattinoz Posts: 1,911member
    robaba said:
    Marvin said:
    Roderikus said:
    OMG 2 years - how hard is it to migrate iMac & MacBook Pro internals into a slightly larger casing ?!
    It depends on what they plan to do with the Mac Pro. It wouldn't make much sense to only put the exact same iMac/MBP chip options in the Mac Pro enclosure so they'd have to manufacture a new chip layout. That takes planning, design, testing, manufacturing, marketing, rollout, it's not a trivial process to make an entirely new chip.


    Not too sure about that last part.  If they use the same M1X SoC designs in a high performance SiP, the on-package power/data fabric should in fact be enough to “glue” the logic units into an effective cpu.  Easier said than done, but that’s what they appear to be attempting.

    but before that tackle this, they’ll want to get the M1X out the door and shook-down for any errata.  The M1X is in production now, so hopefully  the M1Z (?) will be on the bench soon.
    Given they have a Successful 6 core chip already.
    Why wouldn't they test the gluing together 2 chips to make a SIP with the 10-12 core variant. 

    That would get Larger MacBookPro and Lager iMacs into the market and allow them to increment each year with the core September product releases. 
    watto_cobra
  • Reply 35 of 37
    zimmiezimmie Posts: 609member
    zimmie said:
    crowley said:
    michelb76 said:
    imagladry said:
    mike54 said:
    After all the unlimited praise youtubers, tech sites, Apple fanboius, etc gave the M1, I just hope Apple is not taking advantage of this praise, milking as much revenue as they can from it, thereby delaying advancement. Apple does have a bad habit of releasing something great and then sitting on it past its use-by date.
    ah, they produce a new A Series chip every year. My guess is that will be the plan for the M Series, also. Is that "sitting on it past its use-by date?"
    Probably, as the M1 is simply an A14X. 
    It's "simply" something that doesn't exist?  The definition of "simply" must have changed since I was a kid.
    If you look at the A12 versus A12X and A12Z, the changes are almost identical to the changes from the A14 to the M1. CPU goes from two fast cores to four fast cores, GPU goes from four cores to eight cores.

    If you look at the packaging, they're almost identical there, too. The big reason for the M1's limited memory performance and capacity is that it only has space for two RAM packages.

    Thus, the M1 can reasonably be called an A14X.
    So the question is, then (if the rumors are correct and it is a thing), what will the M1X (or A14XX, if you will) be? Just seems like new territory here. Correct me if I’m wrong, but I don’t think there are any true parallels in the A-series history for this “XX” iteration, let alone a third “XXX” iteration as is being speculated on here for a Mac Pro.

    I’ll guess it’s likely they’ll follow the familiar A-series pattern with the M series, so there will be an M1X soon and then an M2 and M2X next year; but that leaves out the Mac Pro. So if there is going to be a third, highest-end, iteration, whether “Z” or something else, kind of seems like it might need to wait for the M2 architecture. 
    This is why I think they'll go with M# Plus for the midrange and M# Pro or M# Max for the top end. Matches iPhone branding they've used for years, and gives a clear "good, better, best" tiering.
    watto_cobra
  • Reply 36 of 37
    robabarobaba Posts: 226member
    mattinoz said:
    robaba said:
    Marvin said:
    Roderikus said:
    OMG 2 years - how hard is it to migrate iMac & MacBook Pro internals into a slightly larger casing ?!
    It depends on what they plan to do with the Mac Pro. It wouldn't make much sense to only put the exact same iMac/MBP chip options in the Mac Pro enclosure so they'd have to manufacture a new chip layout. That takes planning, design, testing, manufacturing, marketing, rollout, it's not a trivial process to make an entirely new chip.


    Not too sure about that last part.  If they use the same M1X SoC designs in a high performance SiP, the on-package power/data fabric should in fact be enough to “glue” the logic units into an effective cpu.  Easier said than done, but that’s what they appear to be attempting.

    but before that tackle this, they’ll want to get the M1X out the door and shook-down for any errata.  The M1X is in production now, so hopefully  the M1Z (?) will be on the bench soon.
    Given they have a Successful 6 core chip already.
    Why wouldn't they test the gluing together 2 chips to make a SIP with the 10-12 core variant. 

    That would get Larger MacBookPro and Lager iMacs into the market and allow them to increment each year with the core September product releases. 
    Depends on several things
    1. whats the optimal high performance to high efficiency ratio in terms of core count?
    2. How many Neural Engine cores do you need?
    3. What level of gpu performance do you need?
    4. What’s the cost per chip.
    5. Can you make a unified chip that you can repackage into other products thereby leveraging your investment into other markets?


    watto_cobra
Sign In or Register to comment.