Intel's chip design, not Apple's choices, reason behind Thunderbolt 3 & RAM issues in new MacBook P

1246710

Comments

  • Reply 61 of 193
    brucemcbrucemc Posts: 1,541member
    Soli said:

    Soli said:
    blastdoor said:
    Well, it's Apple's fault that they keep using Intel processors instead of something based on their wind-storm ARM cores. 
    I would like to see this happen, but this needs to start with the low-end traditional PC offerings, not their Pro-level machines where VMs, Adobe, MS, and other large and complex apps will be used. This needs to happen with a MacBook or Mac mini-like machine where we can have a sub-$800 12" notebook of sub-$500 mini-desktop where you get more performance on their ARM-based SoCs for more users that will most use the default apps and Mac App Store apps.
    And since the choice for the Retina Macbook has been made, isn't this issue sealed by Apple for at least five years?
    Usually Apple is pretty consistent about these things, but they could always throw in a whammy. The 12" MacBook is a neat device but that ultra-low-power Intel Core is crazy expensive for the machine and I think the sales are pretty low so they could pop an ARM chip in there.

    They could also keep the 12" MacBook as is and bring back the MacBook Air, iBook, or something else entirely that runs off of ARM. Marketing is a huge factor so they may not even want to use the Mac name on the HW to help avoid confusion about it not supporting Boot Camp (unless MS offers their ARM-version of Windows) or VMs. The only two things I think we know is that 1) Apple can save hundreds on HW by using their own SoCs, and 2) their SoC already outperform the 12" MacBook and I believe pretty much every WinPC in the range that Apple could then target with a low-cost, entry-level machine based on macOS.
    IF (and I think it is a big "IF" still) Apple does make the decision to utilize ARM for a "macOS / laptop product" they are likely to call it something else (maybe an "Apple Book" given current naming conventions).  It is not a question of technical capabilities, but one of market strategy.  Does Apple want to target the "mid / upper-mid tier" of the laptop market?  They certainly could and it would be a financial success (but Apple clearly does not do things just for that reason).

    The majority of laptop users do not need the power and flexibility that a MBP or iMac x86 provide (VMs, lots of older applications, broadest set of apps & peripherals possible, larger memory requirements, etc).  Apple's native apps + Mac App Store apps + MS Office is likely good enough (to start) for a large set of users.  I am sure Apple could convince some of the top tier app vendors - who already make ARM versions for iPad - to make ARM version of their "PC/Mac" app.  Make a 12" or 13" $799 or $899 machine and Apple would be able to really grow their laptop sales, while maintaining their margins.  This machine would not be under powered - an A11 or A12 derivative SOC could scream for the apps it would support.

    When Apple referred to the A7 as "desktop class 64 bit", was that purely marketing to help position iPhone / iPad as "Post PC", or did they have another plan to slowly take the Mac / laptop line to in-house silicon?  


    Soli
  • Reply 62 of 193
    Mike WuertheleMike Wuerthele Posts: 6,858administrator
    I love the way this story seeks to push the issues people are now commenting on from Apple to Intel. Apple could have introduced a Skylake update to their MB Pro series as soon as the chips were available. Skylake is now over a year old and even the 13 inch chips were introduced in Q1 2016. So I have a question. Can you put USB C and USB A together on the same machine? If the answer is yes then it seems it would have been a much better strategy for Apple to have introduced an old style MB Pro body in the spring of 2016 which included USB C allowing peripheral makers and end users time to start the transition over to the new tech. (This is the way that Apple handled many other transitions away from Firewire, etc). Kaby Lake quad core chips are going to be available in the next few months and this would have meant a much cleaner move over to an all USB C machine in the spring of 2017 with the sexy new touchbar. It seems as if the lunatics are now running the asylum if Apple's product introduction strategy is being decided by the design group rather than when the Intel chips become available. The ones in the middle are us end users who are now looking at the new models thinking why would I invest in a tech that is a year old and which forces me to upgrade all my peripherals or buy a horde of dongles which I have to schlep around with me. For once Appleinsider should question Apple's decisions rather than trying to blame Intel for something which is clearly not their problem!
    You can put USB-A and USB-C on the same machine with the same controller circuitry if they're USB 3.1 generation 1. If you want the USB-C ports to be Generation 2 and the faster speeds, AND Type-A ports on the same machine, that's two chipsets.

    Apple DID make a USB-C machine in 2015/2016. It's called the MacBook. Google's got one too.

    The Surface Book is limited to the same 16GB, but is MS catching heat for that?

    I'm not pinning everything on Intel - but Intel is responsible for a large part of the heat that Apple is getting because of it. The quad-core chips in the 15-inch model didn't ship in capacity until April, and there is no Kaby Lake quad core even sampling yet, well past Intel's guesstimates on when they would ship it.
    watto_cobrakpomroundaboutnow
  • Reply 63 of 193
    The big mistake was Apple not seeing that the intel roadmap was running out of steam, and designing what they need to have innovative new products every 12-14 months.  It's clear, they have to do an x86 chipset.  Going ARM is not a good solution for short term.  

    An AMD acquisition, placed under direction of Apple's current ARM chip team would yield the results we need.  Otherwise the Mac in hamstrung for the foreseeable future.
  • Reply 64 of 193
    brucemc said:
    Soli said:

    Soli said:
    blastdoor said:
    Well, it's Apple's fault that they keep using Intel processors instead of something based on their wind-storm ARM cores. 
    I would like to see this happen, but this needs to start with the low-end traditional PC offerings, not their Pro-level machines where VMs, Adobe, MS, and other large and complex apps will be used. This needs to happen with a MacBook or Mac mini-like machine where we can have a sub-$800 12" notebook of sub-$500 mini-desktop where you get more performance on their ARM-based SoCs for more users that will most use the default apps and Mac App Store apps.
    And since the choice for the Retina Macbook has been made, isn't this issue sealed by Apple for at least five years?
    Usually Apple is pretty consistent about these things, but they could always throw in a whammy. The 12" MacBook is a neat device but that ultra-low-power Intel Core is crazy expensive for the machine and I think the sales are pretty low so they could pop an ARM chip in there.

    They could also keep the 12" MacBook as is and bring back the MacBook Air, iBook, or something else entirely that runs off of ARM. Marketing is a huge factor so they may not even want to use the Mac name on the HW to help avoid confusion about it not supporting Boot Camp (unless MS offers their ARM-version of Windows) or VMs. The only two things I think we know is that 1) Apple can save hundreds on HW by using their own SoCs, and 2) their SoC already outperform the 12" MacBook and I believe pretty much every WinPC in the range that Apple could then target with a low-cost, entry-level machine based on macOS.
    IF (and I think it is a big "IF" still) Apple does make the decision to utilize ARM for a "macOS / laptop product" they are likely to call it something else (maybe an "Apple Book" given current naming conventions).  It is not a question of technical capabilities, but one of market strategy.  Does Apple want to target the "mid / upper-mid tier" of the laptop market?  They certainly could and it would be a financial success (but Apple clearly does not do things just for that reason).
    This. If Apple makes an ARM-based laptop, which I still find unlikely, it would probably be more closely aligned with the iOS architecture than macOS. Making an ARM-based Mac would be a huge mistake, since 1) performance would decrease, and 2) with zero compatibility with existing apps, they'd have the same problem Microsoft had with the early Surface tablets, before they switched them back to Intel, where people were returning the things in record numbers because while they claimed to run Windows, their Windows software didn't work.
    edited October 2016 watto_cobra
  • Reply 65 of 193
    SoliSoli Posts: 10,035member
    brucemc said:
    When Apple referred to the A7 as "desktop class 64 bit", was that purely marketing to help position iPhone / iPad as "Post PC", or did they have another plan to slowly take the Mac / laptop line to in-house silicon?
    I agree with most of your post, but not calling the A7 "desktop class" as purely marketing. Sure, they did it for marketing, just as everything they put a spotlight on is for marketing, but their claim wasn't wrong. There are plenty of AnandTech articles that show that Apple's A-series chips are competitive with the performance found in a huge number of PCs on the market today.
  • Reply 66 of 193
    SoliSoli Posts: 10,035member

    This. If Apple makes an ARM-based laptop, which I still find unlikely, it would probably be more closely aligned with the iOS architecture than macOS. Making an ARM-based Mac would be a huge mistake, since 1) performance would decrease, and 2) with zero compatibility with existing apps, they'd have the same problem Microsoft had with the early Surface tablets, before they switched them back to Intel, where people were returning the things in record numbers because while they claimed to run Windows, their Windows software didn't work.
    It makes no sense to put iOS into a notebook or desktop. Regardless of what they call it, it would be based on macOS. It would have drivers for a trackpad/mouse, keyboard without a virtual keyboard being the primary keyboard input, it would have ports for accessories like a directly connected printer and external display. These are not iOS features.

    And of course they would be compatible with existing apps. Even on the iPhone Pages, for example, is compatible with Pages on macOS. And this isn't just Apple's apps, but innumerable 3rd-party developers make macOS and iOS apps that are compatible with each other. Dropbox, foe example, is only both platforms, as well on Windows and Linux. If and when this does happen you'll see Xcode updated to allow yet another (their 4th?) seamless conversation for developers with relatively little effort, but the apps will exist.
    edited October 2016 watto_cobrabrucemc
  • Reply 67 of 193
    digitoldigitol Posts: 276member
    Ok now we are getting somewhere. So now we are admitting the cpu's are not the best, and in fact a bit dated, and now limiting what a mac can do. Thanks intel. So Apple knows this, and still decides to raise the starting price $500.00?? Oh man. Please don't be taken for a fool. Don't support this laptop. Demand better. 
  • Reply 68 of 193
    Listening to the Intel vs. ARM arguments and was thinking of a story I read last year that said something about the fact that while Intel was still ahead in the performance game, the performance of the A9X was such that there wasn't a lot in it. Is that still the case? And is there a reason why Apple couldn't use emulation as it did when moving from Power PC to Intel chips? With the state of virtualization being accepted today like never before being beholden to a particular chip set seems to be an anathema. Is there a market for lower end machines with slightly less performance but which still run all of the programs you use today? Is it even possible?
  • Reply 69 of 193
    kamilton said:
    The big mistake was Apple not seeing that the intel roadmap was running out of steam, and designing what they need to have innovative new products every 12-14 months.  It's clear, they have to do an x86 chipset.  Going ARM is not a good solution for short term.  

    An AMD acquisition, placed under direction of Apple's current ARM chip team would yield the results we need.  Otherwise the Mac in hamstrung for the foreseeable future.
    The x86 license from AMD cannot be transferred to another company if AMD is bought. Intel can't afford to kill their monopoly.
    williamlondon
  • Reply 70 of 193
    Apple should have altered the form factor to not necessitate the low power versions of these chips - if they tweaked the 'old' design slightly they could have used the currently available more power hungry skylake chips, sacrificed some battery life but had more memory options and and overall more powerful machine - this is a PRO machine, how many pro users are wireless all day?? also, the SSD speeds are to the point where the real world performance is 1-5%, give us user replaceable standard M2 slots that arent a hassle to get to. the new macbook pro doesnt need to compete with the thinness of the regular macbook.
    wiggindigitol
  • Reply 71 of 193
    Soli said:
    It makes no sense to put iOS into a notebook or desktop. Regardless of what they call it, it would be based on macOS. It would have drivers for a trackpad/mouse, keyboard without a virtual keyboard being the primary keyboard input, it would have ports for accessories like a directly connected printer and external display. These are not iOS features.
    Sigh. iOS and macOS Sierra run on top of OS X, which is the same BSDMach Kernel-based OS on both platforms. There are different device drivers, of course, and the iOS interface is based on the CocoaTouch whereas Sierra uses Cocoa, but by and large the same OS is already running on both platforms.

    And you'd have to be particularly obtuse not to believe that a version of macOS isn't happily running on a ARM-based notebook tucked away somewhere in one of Apple's labs.

    Finally, the A10's performance numbers relate directly to the devices (iPhone, iPad) on which it's currently running, which means that it's still throttled back due to power and thermal concerns. I'd love to see what numbers an A10 could crank out with a notebook's heatsinks, fans, and power budget.
    watto_cobratmay
  • Reply 72 of 193
    SoliSoli Posts: 10,035member
    digitol said:
    Ok now we are getting somewhere. So now we are admitting the cpu's are not the best, and in fact a bit dated, and now limiting what a mac can do. Thanks intel. So Apple knows this, and still decides to raise the starting price $500.00?? Oh man. Please don't be taken for a fool. Don't support this laptop. Demand better. 
    1) What's not "better" about a faster CPU, that supports considerably faster I/O, more architecture features, and does it all using less power than Broadwell?

    2) What's not better than this much brighter display with better color, using less power?

    3) What's so necessary about physical PF-keys (do you even know what that means) that you think the Touch Bar is worse?

    4) What's so awful about Touch ID and Apple Pay on a Mac?

    5) What's so bad about 4xUSB-C ports with 5Gibp/s, and 4x USB-C ports with TB3 (or 2xTB3 and 2xTB2) speeds of 40Gibps?

    6) What's worse about no longer having to buy a emplacement or extra power cables from Apple?

    7) What's worse abut having a single, high speed, universal port interface that can be used for I/O charging, I/O data, and display-out to 2x5K or 4x4K displays?
    watto_cobrakpomroundaboutnowksecRayz2016fastasleep
  • Reply 73 of 193
    knowitallknowitall Posts: 1,648member
    kamilton said:
    The big mistake was Apple not seeing that the intel roadmap was running out of steam, and designing what they need to have innovative new products every 12-14 months.  It's clear, they have to do an x86 chipset.  Going ARM is not a good solution for short term.  

    An AMD acquisition, placed under direction of Apple's current ARM chip team would yield the results we need.  Otherwise the Mac in hamstrung for the foreseeable future.
    The x86 license from AMD cannot be transferred to another company if AMD is bought. Intel can't afford to kill their monopoly.
    x86 is dead.
    tmay
  • Reply 74 of 193
    brucemc said:
    Soli said:

    Soli said:
    blastdoor said:
    Well, it's Apple's fault that they keep using Intel processors instead of something based on their wind-storm ARM cores. 
    I would like to see this happen, but this needs to start with the low-end traditional PC offerings, not their Pro-level machines where VMs, Adobe, MS, and other large and complex apps will be used. This needs to happen with a MacBook or Mac mini-like machine where we can have a sub-$800 12" notebook of sub-$500 mini-desktop where you get more performance on their ARM-based SoCs for more users that will most use the default apps and Mac App Store apps.
    And since the choice for the Retina Macbook has been made, isn't this issue sealed by Apple for at least five years?
    Usually Apple is pretty consistent about these things, but they could always throw in a whammy. The 12" MacBook is a neat device but that ultra-low-power Intel Core is crazy expensive for the machine and I think the sales are pretty low so they could pop an ARM chip in there.

    They could also keep the 12" MacBook as is and bring back the MacBook Air, iBook, or something else entirely that runs off of ARM. Marketing is a huge factor so they may not even want to use the Mac name on the HW to help avoid confusion about it not supporting Boot Camp (unless MS offers their ARM-version of Windows) or VMs. The only two things I think we know is that 1) Apple can save hundreds on HW by using their own SoCs, and 2) their SoC already outperform the 12" MacBook and I believe pretty much every WinPC in the range that Apple could then target with a low-cost, entry-level machine based on macOS.
    IF (and I think it is a big "IF" still) Apple does make the decision to utilize ARM for a "macOS / laptop product" they are likely to call it something else (maybe an "Apple Book" given current naming conventions).  It is not a question of technical capabilities, but one of market strategy.  Does Apple want to target the "mid / upper-mid tier" of the laptop market?  They certainly could and it would be a financial success (but Apple clearly does not do things just for that reason).

    The majority of laptop users do not need the power and flexibility that a MBP or iMac x86 provide (VMs, lots of older applications, broadest set of apps & peripherals possible, larger memory requirements, etc).  Apple's native apps + Mac App Store apps + MS Office is likely good enough (to start) for a large set of users.  I am sure Apple could convince some of the top tier app vendors - who already make ARM versions for iPad - to make ARM version of their "PC/Mac" app.  Make a 12" or 13" $799 or $899 machine and Apple would be able to really grow their laptop sales, while maintaining their margins.  This machine would not be under powered - an A11 or A12 derivative SOC could scream for the apps it would support.
    The majority of laptop users do not need the power and flexibility that a MBP or iMac x86 provide is exactly the iPad audience. Apple has made a huge commitment to iPad, deals cut, stocks made, sales high, and Apple has reached even much better marketshare than the laptop. Now it won't kill it for the sake of a crippled ARM notebook. That would be a Chromebook, nothing more.
    When Apple referred to the A7 as "desktop class 64 bit", was that purely marketing to help position iPhone / iPad as "Post PC"?
    Not marketing, a sound and serious policy declaration.

    ... You don't know that yet, but you will also suggest that that ARM notebook have a detachable screen, maybe even a swiveling one...

    edited October 2016
  • Reply 75 of 193
    SoliSoli Posts: 10,035member
    hmlongco said:
    Soli said:
    It makes no sense to put iOS into a notebook or desktop. Regardless of what they call it, it would be based on macOS. It would have drivers for a trackpad/mouse, keyboard without a virtual keyboard being the primary keyboard input, it would have ports for accessories like a directly connected printer and external display. These are not iOS features.
    Sigh.
    It's a different UI. iOS refers to CocoaTouch and the windowless OS, that's why iOS works for the iPad. It would not work for a desktop OS. macOS on ARM does NOT mean you start calling it iOS. The architecture is not the differentiator. If it was, then Mac OS on Motorola chips would not have carried over to PPC, and Mac OS X on PPC would not have carried over to x86.
    edited October 2016
  • Reply 76 of 193
    They made a big mistake with the new MacBooks. They have been making big mistakes since Jobs died. And they have been making poor excuses to cover their big mistakes. Right now they have lost their vision and they will soon lose their following. There is no one at Apple with their fingers on the pulse of the public. Angela A . is a born snob from Cincinnati, the worst kind of snob. She appeals to the 1% and Apple has followed her lead(Wasn't the watch rollout a mess/overpriced)... and will follow her right into the oblivion of... remember when....
  • Reply 77 of 193
    knowitallknowitall Posts: 1,648member
    Listening to the Intel vs. ARM arguments and was thinking of a story I read last year that said something about the fact that while Intel was still ahead in the performance game, the performance of the A9X was such that there wasn't a lot in it. Is that still the case? And is there a reason why Apple couldn't use emulation as it did when moving from Power PC to Intel chips? With the state of virtualization being accepted today like never before being beholden to a particular chip set seems to be an anathema. Is there a market for lower end machines with slightly less performance but which still run all of the programs you use today? Is it even possible?
    A9X runs circles around my 2009 3GHz iMac, and this machine is more than fast enough even for heavy use.
    So, the answer is clear, I think, Apple still has to do some work to connect some extra memory lines to the CPU, to be able to address a terabyte (or so) of RAM, but that's just a few clicks in VHDL.
    Apps can be converted with one click on Xcode, and otherwise be machine (instruction level) translated to ARM and so avoid slow emulation.
    No biggie.
  • Reply 78 of 193
    knowitallknowitall Posts: 1,648member
    Soli said:
    knowitall said:
    Soli said:
    I've never seen an ARM chip that would match the performance of the Mac Pro. Additionally, the Mac Pro is designed to be used with the most processor intensive, complex apps. Even if they had an ARM setup that was double the performance of the Mac Pro you'd still have an issue with app support from Adobe, for one. The only way I see this happening is with marketing to the more basic users that only need the more basic apps. Plus, they can tap into an entirely new customer base with a much cheaper Mac that doesn't reduce their profit margin.
    Pfff, not so. Adobe needs support for OpenCL and OpenGL, both perfectly supported by Apples Ax chips.
    OK, then show me this ARM chip for sale that exceeds the Xeon's used in the Mac Pro.
    Apple doesn't sell components, but I expect the A10X (x 2) will do a pretty good job.
  • Reply 79 of 193
    boeyc15boeyc15 Posts: 986member
    Soli said:
    digitol said:
    Ok now we are getting somewhere. So now we are admitting the cpu's are not the best, and in fact a bit dated, and now limiting what a mac can do. Thanks intel. So Apple knows this, and still decides to raise the starting price $500.00?? Oh man. Please don't be taken for a fool. Don't support this laptop. Demand better. 
    1) What's not "better" about a faster CPU, that supports considerably faster I/O, more architecture features, and does it all using less power than Broadwell?

    2) What's not better than this much brighter display with better color, using less power?

    3) What's so necessary about physical PF-keys (do you even know what that means) that you think the Touch Bar is worse?

    4) What's so awful about Touch ID and Apple Pay on a Mac?

    5) What's so bad about 4xUSB-C ports with 5Gibp/s, and 4x USB-C ports with TB3 (or 2xTB3 and 2xTB2) speeds of 40Gibps?

    6) What's worse about no longer having to buy a emplacement or extra power cables from Apple?

    7) What's worse abut having a single, high speed, universal port interface that can be used for I/O charging, I/O data, and display-out to 2x5K or 4x4K displays?

    Absolutely nothing. although Ill quibble that a legacy 3.0 USB should have been included... there are only a couple gazillion legacy out there that are perfectly functional. And I love the Mag. Magport is rad!

    If I understand this brew-ha-ha correctly there is a class of chips / memory(will call this the PRO class) that could have been added to MacBook Pro that would allow 32GB RAM, but that PRclass of chips was not and has not in been in the previous versions? Simple put --- desktop class chips in the 'portable'. This PRO class of chips would be higher wattage thus require more cooling and battery than the 'new' version chassis could handle.

    And according to the 'PRO's' commenting --- this High RAM configuration usage case is 'becoming very standard' in Professional development world.. and could become a more typical usage cases of 'pro-sumer'(apples present intended market).

    Could all of these things been done in the prior chassis? 

    In summary, in my ass-hat opinion -- If Apple kept a version of the old basic chassis (tweaked it) and did all above and had two internal board/battery configs(yes it cost money) --- 16GB board(medium duty chips) and 32 gb Board(heavy duty chips) it would be darn closer to meeting 100% of it user community's needs, including an absolutely kick-arse machine(albeit lower batter life). The down side-- two configs to manage, and ohh yes... Apples obsession with weight and thinness would have to be sacrificed a bit.

    I understand Apples design for 80%-90% of users, new tech for old etc, but in my opinion, they used to design and sell hardware that was 'optioned' to be much closer to 100% of its users needs and the 'hardware' features/ports were more consistent and allowed transition better. Now its like a mish-mash of features/ports across devises with one size fits all ports and processors for each device (Sounds like a 'supply guy' running things at Apple! ).




    edited October 2016
  • Reply 80 of 193
    VSzulcVSzulc Posts: 32unconfirmed, member
    This is such nonsense, and the saddest attempt at shilling for a corporation, I've seen in a long time.

    Here are the facts:

    FACT: Intel Skylake CPUs support up to 64GB of RAM. Mobile CPUs included.

    FACT: LPDDR3 comes in 32GB SODIMMs these days. Samsung is one manufacturer.

    FACT: If Apple had the best interest of their pro customers in mind, rather than how thin their MBP is, they could have installed one of these, or used 2 16 GB SODIMMs

    Or even better, put 8GB RAM on the motherboard, and left an expansion port for future use. They could also have used DDR3L RAM, for those customers who don't mind taking a small hit on battery life.

    FACT: There are several ways they could have accomplished this. There are plenty of 15inch thin and light Windows laptops with 32GB RAM at lower prices than the MBP, and at a similar size/weight. Why they didn't is a good question, but the answers they've given are nonsense.

    Another good question is why Appleinsider is trying to defend Apples poor decision, rather than look out for the interest if Apples customers and their readers.
    edited October 2016
Sign In or Register to comment.