Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
2018 i9 MacBook Pro throttling discussion [merged]
- Thread starter M.Rizk
- Start date
-
- Tags
- macbook performance pro
- Sort by reaction score
An expensive and embarrassing mistake by Apple, but As Oscar Wilde said: "The only thing worse than being talked about is not being talked about."
Ok so I decided to open mine after seeing everybody's results 
If problems do come up in the coming 14 days I'll still return it.
I ran Cinebench 5 consecutive times on a clean account with no wifi and ambient temp of 25 degrees Celsius.
Run Score
1 1096
2 1072
3 1050
4 1056
5 1093
So I guess that's pretty good right? Temperature went up to 100 degrees on turboboost but then got down to 92/93 right after, with speed staying between 3.2 and 3.3Ghz.
The machine did get hot, warming up the service below it quite a lot, same with the keyboard, but so did my MBP 2014. Looking forward to seeing everybody's real life benchmarks now. Personally I'd be interested to see a benchmark with a eGPU attached so the internal GPU's don't use energy. I think it may be able to score higher then, but not sure?
Now that its syncing everything from iCloud it seems to hover around 4.2Ghz at 82 degrees Celsius, for whatever that's worth.
Edit: Results come from my i9, 32gb, 1tb
If problems do come up in the coming 14 days I'll still return it.I ran Cinebench 5 consecutive times on a clean account with no wifi and ambient temp of 25 degrees Celsius.
Run Score
1 1096
2 1072
3 1050
4 1056
5 1093
So I guess that's pretty good right? Temperature went up to 100 degrees on turboboost but then got down to 92/93 right after, with speed staying between 3.2 and 3.3Ghz.
The machine did get hot, warming up the service below it quite a lot, same with the keyboard, but so did my MBP 2014. Looking forward to seeing everybody's real life benchmarks now. Personally I'd be interested to see a benchmark with a eGPU attached so the internal GPU's don't use energy. I think it may be able to score higher then, but not sure?
Now that its syncing everything from iCloud it seems to hover around 4.2Ghz at 82 degrees Celsius, for whatever that's worth.
Edit: Results come from my i9, 32gb, 1tb
Last edited:
Ok so I decided to open mine after seeing everybody's results
I ran Cinebench 5 consecutive times on a clean account with no wifi and ambient temp of 25 degrees Celsius.
Run Score
1 1096
2 1072
3 1050
4 1056
5 1093
So I guess that's pretty good right? Temperature went up to 100 degrees on turboboost but then got down to 92/93 right after, with speed staying between 3.2 and 3.3Ghz.
The machine did get hot, warming up the service below it quite a lot, same with the keyboard, but so did my MBP 2014. Looking forward to seeing everybody's real life benchmarks now. Personally I'd be interested to see a benchmark with a eGPU attached so the internal GPU's don't use energy. I think it may be able to score higher then, but not sure?
Now that its syncing everything from iCloud it seems to hover around 4.2Ghz at 82 degrees Celsius, for whatever that's worth.
Is this coming from an i9?
Yes sorry i9 32gb 1tb. I had a discussion in a different thread that on these kind of tests with a sustained load, we kind of can't expect i9 to score higher than the 2.2 and 2.6 as they all can sustain speeds up to 3.2ghz in this case.
I guess the i9 only makes sense if you have a workflow that uses short bursts to reach the 4.3 and higher Turbo Boost for a (couple of) second(s).
I’m curious what kind of workflows would benefit from something like that?
I do a lot of rendering in After Effects of short (5 minutes or less) HD scenes so I’m thinking it might not benefit my scenario.
Not sure, which is why I'm still completely open to bring this back if an error occurs or if all the benchmarks show no benefit to the i9. My workload is pretty diverse going from designing, to programming to rendering, so I'll be looking at all those results for the next 13 days to base the decision on. If there are no benefits I can use the €300 to put towards an eGPU case.
What we need now is some proper testing of the 2.6 i7 vs the 2.9 i9. I want to see if they basically operate at the same turbo boost frequencies during both short term and sustained loads. I can save a bunch of money if the i5 performs close to identical to the i7
The base i7 2.2 ghz model seems to be the sweet spot, since all 3 cpus meet at about the same turbo frequency when pushed. As seen till now, more cache on i9 doesn't affect that much.
Last edited:
This to me proves that all 3 of these processors are going to be about the same due to the heat output and necessity to throttle when pushed. Very interesting power limiting discussion
Could you please do testing of many tracks, many plugins, and LOW (LOWEST) buffer size setting and reveal how it performs without your own fan management in terms of heat, clicks & pops, and fan speed/noise during that process? Thx
Bootcamp/win10 and vm/win10 stats would also be interesting. Apple may still have to deploy a fix for those who switch OSes a lot (like me). Has any brave soul attempted this?
I'm thinking about the automatic test runners we have. When new source code files are saved, the test runners can kick in and run tests for a few seconds and give immediate feedback on pass/fail.
In this kind of workflow, you'd see short bursts on the order of several seconds, over and over again with pauses in-between for human coders to catch up.
How likely do we get firmware upgrade so Bootcamp users can benefit the fix?
BootCamp runs within macOS so it will have the firmware already downloaded into the ROM's. Are you asking how to get a straight Windows MacBook Pro to get the firmware update? You'll need to boot the system up under macOS and once its installed into the ROM you're good to go! Nothing special otherwise.
Hope this post is OK. If not, sorry mods, I will remove it. If any members are willing to run Unreal Engine tests on their new machines with the patch applied, please post here - https://forums.macrumors.com/threads/2018-macbook-pro-unreal-engine-light-bake-test.2129380/
Bootcamp doesn't "run" in macOS. That's the whole point. VMs like Parallels and Fusion do that. As some are indicating already here, a "fix" may not be ready for Bootcamp users. But I'm sure there's more to come (and associated benchmarks) on that front.
Somebody mentioned that there are lots of programs to twist under Windows. Do we need to change the settings whenever we reboot into Windows?
[doublepost=1532539035][/doublepost]Why they call it firmware upgrade? I think it is just a software upgrade under Mac OS running the latest OS only.
https://www.pcmag.com/news/362510/apple-macbook-pro-benchmarked-fast-but-surprise-youll
[doublepost=1532539035][/doublepost]Why they call it firmware upgrade? I think it is just a software upgrade under Mac OS running the latest OS only.
https://www.pcmag.com/news/362510/apple-macbook-pro-benchmarked-fast-but-surprise-youll
I used Intel Xtu, but you need to set the values after each restart (at least I think so?)