Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
So your 2017 hits 100c at one point? That in a weird sort of way is a bit encouraging. Below's mine and


I seem to be getting better numbers from my 2.2 base model then your 2.6 model :eek:

I'm using Macs Fan Control, but even without that, I'm seeing CPU only numbers in the 920 range. It spikes right up at 100c and then starts banging around the 90c range until it recedes back to something more respectable.

View attachment 771714

The high temperatures people are seeing are normal for an MBP. Apple throttle by temperature, not TDP as most others do. Typically MBP applies thermal throttling at around 100C. My 2015 throttles at a little under 100 degrees. When the cooling system manages to lower the temp a little I get a little boost, then a temp increase, slight throttle, temp decrease and so on.

Of course that does not take away from the fact that i9’s look to be go getting hot very fast and the cooling struggles to keep up and sub-base speeds are seen. In other words I’m not saying this means they are cooling well, just that it’s entirely normal/expected to see an MBP to get to 100 at full tilt.

Edit: here’s a credible source for that claim https://www.notebookcheck.net/Apple...e-performance-with-a-few-clicks.317552.0.html

Edit 2: Well, actually after reading the article I linked (I just searched for something supporting what I was saying about temperature throttling) it seems Apple’s strategy of throttling by temp instead of TDP is the root of the issue. Notebookcheck used utilities to limit TDP and the performance was much better. Light at the end of the tunnel?

“notebookcheck” said:
Verdict
Apple's philosophy of removing all consumption limitations is clearly counterproductive for the current 2018 MacBook Pro systems. Even very short load periods of ~30 seconds result in massive clock fluctuations, which will affect the performance. We recommend the manual adjustment of the CPU consumption for both model, but the 15-inch MBP in particular. You still get the maximum Turbo Boost when a single core is stressed, and the performance is better and especially steadier under maximum load. We think Apple's engineers should have figured this out and a simple software update would solve the issue, but we know that the manufacturer from Cupertino does not like to admit these things (also see keyboard problems).
 
Last edited:
Your CPU was only ever under a 50% load, so if it had throttled under such a light load that would be … very bad.

Do something that puts all 6 cores under heavy load for more than a minute or two and see what happens.
Like what would you suggest? I just did a 10 minute final cut 4k render.... My regular workflow, and it held up pretty consistently. I'm just demonstrating 'real world' usage (I do video full time).
 
Like what would you suggest? I just did a 10 minute final cut 4k render.... My regular workflow, and it held up pretty consistently. I'm just demonstrating 'real world' usage (I do video full time).

If that’s your regular workflow, that’s great. No need to artificially force a throttle with an unrealistic scenario.

I will say, though, that I have 2 machines, one a 2016 MacBook Pro, and one a gaming notebook with a 7700HQ, and I can’t get either of those machines to throttle no matter what I do, even things like Prime95 torture test.

Is it normal for rendering to only use 50% of the CPU? I thought it would max it out? Isn’t that the whole point of having a fast CPU with many cores?
 
  • Like
Reactions: uecker87
It seems to me that this generation of i7 CPU's temperature will suddenly gets very high (under load) and then cool down quite quickly, just like it suddenly produces a lot of heat and then dissipated by the cooling system. I don't feel that on my 2016 MBP.
 
I have been running cinebench for more than 20 rounds and it is still stable so far. :eek: .


Hey nice. What are your fans (fan?) doing? I'm hoping to not use fan control. Are yours at idle?

I use volta to undervolt -150mv and set power limit to 45W. The 2.2ghz based i7 locked in at 3ghz and achieves around 950 cinebench on average even after 10 runs. If I do not set the power limit, the clocks keep jumping all over the place and only get about 820 cinebench score.
https://www.flickr.com/photos/157256909@N08/shares/R495Cc
R495Cc
 
Prime 95 torture test... about 12 mins in:
Screen Shot 2018-07-21 at 02.12.05.png

[doublepost=1532135694][/doublepost]
I’d love to compare speeds with your machine. Do you have After Effects to compare speeds with? Or a FCP X Benchmark Project?
I'll try and grab something... will be tomorrow though as I'm in the UK and its 2:14am!
 
Unrelated to cpu, but how much would it affect temps if I were to upgrade from 555X to 560X? I am going with the 2.2ghz i7 but not sure about the gpu. It is small enough to warrant the upgrade? I plan on using it mainly and heavily for music production in Logic Pro and Pro Tools.
 
  • Like
Reactions: Petetastic
Unrelated to cpu, but how much would it affect temps if I were to upgrade from 555X to 560X? I am going with the 2.2ghz i7 but not sure about the gpu. It is small enough to warrant the upgrade? I plan on using it mainly and heavily for music production in Logic Pro and Pro Tools.
It wouldn't change temps. But that's because it also effectively doesn't change performance. Don't get that upgrade.
 
Unrelated to cpu, but how much would it affect temps if I were to upgrade from 555X to 560X? I am going with the 2.2ghz i7 but not sure about the gpu. It is small enough to warrant the upgrade? I plan on using it mainly and heavily for music production in Logic Pro and Pro Tools.

I’m in the exact same boat as you.
For the little price upgrade I might get the 560 even though we probably will never know we have it.

Thinking about the i7 2.2 with 32GB ram. Should handle all the music projects easily.
 
  • Like
Reactions: Petetastic
https://www.computerworld.com/artic...hat-macbook-pro-core-i9-throttling-story.html


What about the tests?
While Lee has gained attention with his claims, it seems possible they may be overblown. Not only does it seem possible Adobe has yet to upgrade Premiere to get best results from Apple’s new Macs, but even the Cinebench test Lee used to clock the speeds may be inaccurate.

Perhaps the software used in the test is to blame?
YouTuber, Jonathan Morrison is running his own series of tests on one of the latest MacBook Pros running a Core i9 chip.

His tests show Final Cut Pro X delivers astonishingly better performance when running a series of relatively common pro workflows on one of the new Macs than you can expect when using Adobe Premiere Pro for the same tasks.


...


Perhaps the test suite needs a patch?
It is possible there’s a problem in the test software used in Lee’s tests.

Dieter Bohn took a moment to ask Cinebench if the test software was accurate on the new Mac.

The response?

Cinebench said its application: “Has not been upgraded to measure the performance in a meaningful way,” adding that, “the development team is aware and will be addressing this in the future.”
 
  • Like
Reactions: Petetastic
https://www.computerworld.com/artic...hat-macbook-pro-core-i9-throttling-story.html


What about the tests?
While Lee has gained attention with his claims, it seems possible they may be overblown. Not only does it seem possible Adobe has yet to upgrade Premiere to get best results from Apple’s new Macs, but even the Cinebench test Lee used to clock the speeds may be inaccurate.

Perhaps the software used in the test is to blame?
YouTuber, Jonathan Morrison is running his own series of tests on one of the latest MacBook Pros running a Core i9 chip.

His tests show Final Cut Pro X delivers astonishingly better performance when running a series of relatively common pro workflows on one of the new Macs than you can expect when using Adobe Premiere Pro for the same tasks.


...


Perhaps the test suite needs a patch?
It is possible there’s a problem in the test software used in Lee’s tests.

Dieter Bohn took a moment to ask Cinebench if the test software was accurate on the new Mac.

The response?

Cinebench said its application: “Has not been upgraded to measure the performance in a meaningful way,” adding that, “the development team is aware and will be addressing this in the future.”


The Apple sales I talked to said that there is nothing wrong with the i9 MBP. He put all the blames on we users not using up-to-date software for benchmarking!
 
Cinebench said its application: “Has not been upgraded to measure the performance in a meaningful way,” adding that, “the development team is aware and will be addressing this in the future.”
What could they possibly do to Cinebench to allow it to "measure the performance in a meaningful way"?
 
What could they possibly do to Cinebench to allow it to "measure the performance in a meaningful way"?

Ostensibly there is something or else why would they issue that statement? That’s the puzzler and why I felt that article significant enough to post.
 
Ostensibly there is something or else why would they issue that statement? That’s the puzzler and why I felt that article significant enough to post.
It's just an excuse. Cinebench has been used to benchmark for who knows how long and this is the first time I've ever heard it might need modification. I suspect if I were to run the same software / tests on my Mac Pro I wouldn't see any thermal throttling.
 
It's just an excuse. Cinebench has been used to benchmark for who knows how long and this is the first time I've ever heard it might need modification. I suspect if I were to run the same software / tests on my Mac Pro I wouldn't see any thermal throttling.

There's certainly non on of my W10 notebook with the 8750H, as ever with Apple Smoke & Mirrors. It would litterly kill them to present an honest answer.

Most of all it's plain insulting...

Q-6
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.