Tested: Thermal conditions in the 2018 i9 MacBook Pro dramatically hampering performance

245678

Comments

  • Reply 21 of 158
    DuhSesameDuhSesame Posts: 281member

    Apple has been form over function for some time. It's unfortunate, their quest for sleek and thin, rather than functional. Not only performance, but keyboard ergonomics (and reliability) are at issue with the 2016 and newer models. The same has of course been true of the current Mac Pro - too much form over function, and customers started to turn up their noses. Apple needs to reel in their designers and allow the engineers to build better, more functional machines. The quest for thin and small needs to stop, and reverse course a bit.
    Like the billion times I said, only gaming laptops can keep an i9 under control.  MacBook Pros aren't that bad when compared to others under 25mm.
    edited July 18 orthorimAvieshekmacxpress
  • Reply 22 of 158
    Is the i7 throttled as well? How does i7 performance compare to the spec? How does i7 throttled compare to i9 throttled?
    Avieshek
  • Reply 23 of 158
    DuhSesameDuhSesame Posts: 281member
    Is the i7 throttled as well? How does i7 performance compare to the spec? How does i7 throttled compare to i9 throttled?
    We have to wait until there's other reviews.  But don't expect to be much better.
    Avieshek
  • Reply 24 of 158
    You’re doing it wrong.  Offices are supposed to be well chilled.

    Kind of sad the i9 is a waste of money...

    I’m sure the iMac Pro also has thermal issues.  All of Apple’s designers/QC testers must be assigned to iPhone & IPad...
  • Reply 25 of 158
    zeudevzeudev Posts: 1member
    Hi Mike,

    Do you have any interest in running another benchmark with just the cpu engaged?  For some of the software developers in your audience, they will only care how well it performs when only the cpu is kicking out heat.
  • Reply 26 of 158
    DuhSesameDuhSesame Posts: 281member
    You’re doing it wrong.  Offices are supposed to be well chilled.

    Kind of sad the i9 is a waste of money...

    I’m sure the iMac Pro also has thermal issues.  All of Apple’s designers/QC testers must be assigned to iPhone & IPad...
    Well, short-time burst, so not all waste.  iMac Pro probably can keep the 18-core under control, at least not a terrific drop.

    I think that’s the problem with Intel, which they did not really build their mobile processors for anything under 25mm.  If that’s the case, simply let it run slower so people will purchase those really huge ones.

    Intel screwed up thermal efficiency quiet a few times since Pentium 4 (I’m sure some of you rememberers that). 

    Prescott because the architecture, Haswell because of FIVR, and Coffee Lake because it can’t bring 10nm in.
    edited July 18 Avieshek
  • Reply 27 of 158
    Mike WuertheleMike Wuerthele Posts: 3,121administrator
    zeudev said:
    Hi Mike,

    Do you have any interest in running another benchmark with just the cpu engaged?  For some of the software developers in your audience, they will only care how well it performs when only the cpu is kicking out heat.
    We're working on it.
    orthorimcaladaniancgWerksjdwaknabiAvieshek
  • Reply 28 of 158
    mdriftmeyermdriftmeyer Posts: 7,127member
    Time for Apple to go AMD with Zen+ and Zen2, etc. The TDP and performance metrics are out to show that AMD is now back on top.
    blastdoorAvieshekwilliamlondon
  • Reply 29 of 158
    DuhSesameDuhSesame Posts: 281member
    Time for Apple to go AMD with Zen+ and Zen2, etc. The TDP and performance metrics are out to show that AMD is now back on top.
    That's a really good alternative.  I think their high-end mobile processors will be ready in next year or so?  Unless Apple decides to shift to its own.
    edited July 18 Avieshek
  • Reply 30 of 158
    orthorimorthorim Posts: 141member
    blastdoor said:
    I have a 2017 15” MBP. I like it. I wouldn’t want them to change the form factor.

    There are two fundamental problems here from my point of view:

    1. Apple should not offer the i9 as an option if it has no real advantage. A big part of the advantage of Apple is that they don’t dump design decisions on the user. 

    2. Apple needs to offer better pro desktop options. The iMac pro is ok but we need a real pro desktop. If there were a real pro desktop option then people would be more willing to accept the compromises inherent in an elegantly designed laptop 


    I guess blame it all on Intel. That's what Apple is doing - Intel promised lower power chips, and didn't deliver.

    If they did not offer the i9, Dave Lee and pretty much everyone else on the internet would endlessly whine about it. As it is, shipping the i9 was the right decision even though I'm super disappointed with this and won't get one as planned now.

    It's possible Intel will fix this by providing better chips to Apple, or Apple can do some magic with fan management - it's still possible there are bugs in the fan management software and it's just throttling more than it needs to be.

    @AppleInsider - I would love for you to do a processor test that doesn't use the GPU as well. 

    Because if you use the GPU, there's a second massive heat source in the system. If you run without GPU, you should see less throttling?! I am not sure Adobe Premiere and Cinebench are using the GPU, but the article seems to imply at lease Premiere does. 

    In my own tests trying to play a game on my MBP, just running the GPU would nearly overheat the system, didn't even need a lot of CPU power for that.
    caladanian
  • Reply 31 of 158
    cpsrocpsro Posts: 2,350member
    FWIW: I repeatedly got GB4 single-core scores of 504X and multi-core scores of 224XX on a 2.2 GHz model on display in an Apple Store. If GB4 scaled linearly with clock speed, one would predict a multicore score of 29600 for the Core i9 model. This isn't a completely unreasonable expectation because the i9 has 12 MB L3 cache vs. only 9 MB in the 6-core i7 models. Due to throttling and the fact that main memory is the same speed across all models, the best i9-8950HK benchmarks don't even come close to 29600. Even PCs with faster 2666 memory aren't close to 29600.
    d_2Avieshek
  • Reply 32 of 158
    vadimyuryevvadimyuryev Posts: 135member, editor
    Is the i7 throttled as well? How does i7 performance compare to the spec? How does i7 throttled compare to i9 throttled?
    Watch the video.

    The i7 (advertised at 2.2GHz base & 4.1GHz Turbo Boost) stayed at around 2.3 - 2.6GHz average clock speed during back-to-back CPU tests in Cinebench R15.
    Reached a peak of 3.8GHz when the test started.

    The i9 (advertised at 2.9GHz base & 4.8GHz Turbo Boost) was at around 2.4 - 2.8GHz average clock speed.
    Reached a peak of 4.2GHz when the test started.
    d_2bayridge8829lamboaudi4
  • Reply 33 of 158
    vadimyuryevvadimyuryev Posts: 135member, editor
    orthorim said:
    blastdoor said:
    I have a 2017 15” MBP. I like it. I wouldn’t want them to change the form factor.

    There are two fundamental problems here from my point of view:

    1. Apple should not offer the i9 as an option if it has no real advantage. A big part of the advantage of Apple is that they don’t dump design decisions on the user. 

    2. Apple needs to offer better pro desktop options. The iMac pro is ok but we need a real pro desktop. If there were a real pro desktop option then people would be more willing to accept the compromises inherent in an elegantly designed laptop 


    I guess blame it all on Intel. That's what Apple is doing - Intel promised lower power chips, and didn't deliver.

    If they did not offer the i9, Dave Lee and pretty much everyone else on the internet would endlessly whine about it. As it is, shipping the i9 was the right decision even though I'm super disappointed with this and won't get one as planned now.

    It's possible Intel will fix this by providing better chips to Apple, or Apple can do some magic with fan management - it's still possible there are bugs in the fan management software and it's just throttling more than it needs to be.

    @AppleInsider - I would love for you to do a processor test that doesn't use the GPU as well. 

    Because if you use the GPU, there's a second massive heat source in the system. If you run without GPU, you should see less throttling?! I am not sure Adobe Premiere and Cinebench are using the GPU, but the article seems to imply at lease Premiere does. 

    In my own tests trying to play a game on my MBP, just running the GPU would nearly overheat the system, didn't even need a lot of CPU power for that.
    The test in the video doesn't use the GPU.
    orthorimAvieshek
  • Reply 34 of 158
    rufworkrufwork Posts: 126member
    Maybe I scanned something wrong, but the difference between your 2018 and 2017 numbers seems off. You have 24326 for 2018 i9 and 22179 for 2017 i7. MacRumors has 22439 for i9 and 15548 for i7.

    You're going from four to six cores. Seems like your increase for 6 core should be more than 9.6% more than four. I'd guess those i7 numbers in the video are wrong.

    What am I missing?
    edited July 18 Avieshek
  • Reply 35 of 158
    You’re doing it wrong.  Offices are supposed to be well chilled.

    Kind of sad the i9 is a waste of money...

    I’m sure the iMac Pro also has thermal issues.  All of Apple’s designers/QC testers must be assigned to iPhone & IPad...
    The iMac Pro does not(!) have thermal issues. :) Other iMacs have (different heat flow).
    Avieshek
  • Reply 36 of 158
    Mike WuertheleMike Wuerthele Posts: 3,121administrator
    rufwork said:
    Maybe I scanned something wrong, but the difference between your 2018 and 2017 numbers seems off. You have 24326 for 2018 i9 and 22179 for 2017 i7. MacRumors has 22439 for i9 and 15548 for i7.

    You're going from four to six cores. Seems like your increase for 6 core should be more than 9.6% more than four. I'd guess those i7 numbers in the video are wrong.

    What am I missing?
    Are you talking about the article or the video? The article doesn't address the 2017 benchmarks at all, really.
  • Reply 37 of 158
    orthorimorthorim Posts: 141member
    orthorim said:
    blastdoor said:
    I have a 2017 15” MBP. I like it. I wouldn’t want them to change the form factor.

    There are two fundamental problems here from my point of view:

    1. Apple should not offer the i9 as an option if it has no real advantage. A big part of the advantage of Apple is that they don’t dump design decisions on the user. 

    2. Apple needs to offer better pro desktop options. The iMac pro is ok but we need a real pro desktop. If there were a real pro desktop option then people would be more willing to accept the compromises inherent in an elegantly designed laptop 


    I guess blame it all on Intel. That's what Apple is doing - Intel promised lower power chips, and didn't deliver.

    If they did not offer the i9, Dave Lee and pretty much everyone else on the internet would endlessly whine about it. As it is, shipping the i9 was the right decision even though I'm super disappointed with this and won't get one as planned now.

    It's possible Intel will fix this by providing better chips to Apple, or Apple can do some magic with fan management - it's still possible there are bugs in the fan management software and it's just throttling more than it needs to be.

    @AppleInsider - I would love for you to do a processor test that doesn't use the GPU as well. 

    Because if you use the GPU, there's a second massive heat source in the system. If you run without GPU, you should see less throttling?! I am not sure Adobe Premiere and Cinebench are using the GPU, but the article seems to imply at lease Premiere does. 

    In my own tests trying to play a game on my MBP, just running the GPU would nearly overheat the system, didn't even need a lot of CPU power for that.
    The test in the video doesn't use the GPU.
    Oh. Duh. Should have watched the video. 

    So I guess the i9 is like a Ferrari in NYC traffic right now.... damn. I hope they'll fix / address it, I'll give them a few days.
  • Reply 38 of 158
    Is the i7 throttled as well? How does i7 performance compare to the spec? How does i7 throttled compare to i9 throttled?
    Watch the video.

    The i7 (advertised at 2.2GHz base & 4.1GHz Turbo Boost) stayed at around 2.3 - 2.6GHz average clock speed during back-to-back CPU tests in Cinebench R15.
    Reached a peak of 3.8GHz when the test started.

    The i9 (advertised at 2.9GHz base & 4.8GHz Turbo Boost) was at around 2.4 - 2.8GHz average clock speed.
    Reached a peak of 4.2GHz when the test started.
    Thanks. Hadn’t watched the video yet. I meant to ask about the 2.6 i7 vs the 2.2 i7. I already have an order shipping to me for i9. Wondering if I should return and go for base i7 2.2 or i7 2.6. If thermals are going to hold me back, then perhaps the base i7 is fine because the others can’t reach their potential. I’m coming from an i5 mid-2014 13” Pro, so any of these will be an upgrade. 
  • Reply 39 of 158
    jdb8167jdb8167 Posts: 105member
    anome said:

    Cheap gags aside, am I to understand we're not seeing similar issues with the quad-core 13"? Or the hex-core i7? Or is it just that everyone bought the i9 because they wanted MOAR POWER!!!, and so we don't have anyone testing the lower tiers?

    I did an informal test with my new 13” MBP with an i7. I have a command line tool that runs multiple copies of the OpenSSL speed test that maxes out all 8 threads. I let it run for 10 minutes until the fans were maxed out and then ran GB4. I saw a minor 1% to 2.5% slow down. It doesn’t look like it is throttling much. I plan to do more testing this weekend. The core i7 13” MBP is a beast. I’m getting 5400 single core and 19100 multi core.
    edited July 18 cgWerksanome
  • Reply 40 of 158
    Mike WuertheleMike Wuerthele Posts: 3,121administrator
    You’re doing it wrong.  Offices are supposed to be well chilled.

    Kind of sad the i9 is a waste of money...

    I’m sure the iMac Pro also has thermal issues.  All of Apple’s designers/QC testers must be assigned to iPhone & IPad...
    It does not.
    macxpresscornchip
Sign In or Register to comment.