Probably I saw a video but they are sorta known for doing that. Conservative fan speeds that is
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
It also just hit me that if scores aren’t showing decreased performance by a large margin then that power gadget may simply be incorrectly reading the processor.
I re-ran it again and downloaded "Macs Fan Control 1.4.10". At about 20 seconds is when I could start hearing the fans obviously (fans @ ~3100 rpm) and then they spun up over about 30 seconds to ~5100 rpm and didn't go much faster. The app reports a maximum fan speed of 5927 rpm on one fan and 5489 rpm on the other fan. Interesting that.
The fans are not obnoxiously loud at all even when they hit 5200. It's surprisingly quiet actually. I have no idea how to accurately measure that for you though.
Edit: https://i.imgur.com/FeVFgHU.png
I'm not sure how to explain the core vs. package temperature discrepancy though, or what the difference practically is.
Edit: Scratch a lot of those specific numbers. I think Mac Fan Control doesn't update values very frequently so they were probably lagging a couple hundred ms. Comments on noise level stand though.
Last edited:
Ummm, those of you that expect full throttle from a high end mobile processor over extended periods need to just go buy a desktop. All laptops have similar thermal limitations with these kinds of processors.
What exactly was the test methodology used for these vids? The one is ok as he explains he has made the config 'exactly' the same. The other just says it's his co-workers laptop.
Anyway, I trust this review far more than some random YouTube guy looking for eyeballs...
Edit - The author is a NASA engineer using his laptop for simulations. He compares it with a 2017 iMac, a 2018 iMac Pro and a 2016 Macbook pro. It compares very favorably up to the number of cores it has.
What exactly was the test methodology used for these vids? The one is ok as he explains he has made the config 'exactly' the same. The other just says it's his co-workers laptop.
Anyway, I trust this review far more than some random YouTube guy looking for eyeballs...
Edit - The author is a NASA engineer using his laptop for simulations. He compares it with a 2017 iMac, a 2018 iMac Pro and a 2016 Macbook pro. It compares very favorably up to the number of cores it has.
Last edited:
Maybe, maybe not.
Instead of the Power Gadget (which I believe to be accurate) we need some straight timed tests of multiple instances of BruceX, or whatever the project may be, across the range of CPU options. Get them running full tilt and assess the results based on time. Power Gadget can support the findings.
I’m leaning very much toward the i9 hitting the thermal cap and being limited due to the chassis design and cooling system.
Let’s see some timed tests to disprove this and show the i9 as worth the cash.
Instead of the Power Gadget (which I believe to be accurate) we need some straight timed tests of multiple instances of BruceX, or whatever the project may be, across the range of CPU options. Get them running full tilt and assess the results based on time. Power Gadget can support the findings.
I’m leaning very much toward the i9 hitting the thermal cap and being limited due to the chassis design and cooling system.
Let’s see some timed tests to disprove this and show the i9 as worth the cash.
The i9 is a hot CPU with a higher clock speed and thermal profile. It's been dropped into a laptop with the same thermal limitations as the previous quad core based generation. It's going to throttle. You might be better off going for the i7 which has a lower thermal profile and clock speed, generates less heat and is therefore able to run closer to it's capacity for longer than the i9.
The Coffee Lake i9 is a CPU with about a 40% increased thermal profile over the last generation i7-7700. The throttling experienced here isn't a fault and definitely isn't limited to Apple. It's simply the result of dropping a much more powerful CPU which generates far more heat into a laptop with the same thermal capacity as before.
The Coffee Lake i9 is a CPU with about a 40% increased thermal profile over the last generation i7-7700. The throttling experienced here isn't a fault and definitely isn't limited to Apple. It's simply the result of dropping a much more powerful CPU which generates far more heat into a laptop with the same thermal capacity as before.
Upgrading to the max CPU option always has very little return. Save the money for RAM, SSD, graphics and accessories.
Because somebody asked me on reddit I ran a test with the fans at maximum (using Mac Fan Control) and Prime 95 again.
https://imgur.com/a/NQ1b7G2
Very quickly the CPU frequency falls to ~2.5 Ghz and stays there. What's weird in this case is that the CPU temperatures seem... fine. Toasty, but fine, should have more overhead.
https://imgur.com/a/NQ1b7G2
Very quickly the CPU frequency falls to ~2.5 Ghz and stays there. What's weird in this case is that the CPU temperatures seem... fine. Toasty, but fine, should have more overhead.
... Also if anybody has any workloads they want to walk me through let me know, I'm still awake for another hour or so.
[doublepost=1531891875][/doublepost]
The CPU is running 400Mhz BELOW base clock, after literal seconds of load, with the fans manually set to MAX.
How is this a "Boom!?"
The fans are okay at the maximum the normal fan curve goes to (5200ish RPM on the faster one), but are loud at MAX.
[doublepost=1531891875][/doublepost]
The CPU is running 400Mhz BELOW base clock, after literal seconds of load, with the fans manually set to MAX.
How is this a "Boom!?"
The fans are okay at the maximum the normal fan curve goes to (5200ish RPM on the faster one), but are loud at MAX.
Bear in mind you have a toasty GPU sharing the same cooling solution even if it's segmented, the system is therefore juggling between keeping overhead free for the GPU and a high-end CPU. I'm guessing the thermal profile will lock the CPU once it throttles for an extended period to keep temperatures under control. Apple will also want to avoid having the fan spin up to maximum speeds for prolonged periods if at all possible.
There are examples of laptops with high-end CPU's which lock the CPU to less than 1GHz once throttling occurs.
It's a boom in that Apple clearly could have let the CPU clock higher since the temps are stable at 80C in your test with the throttle only down 400mhz. At the very least it should be able to maintain base clock at what I would guess would be 90C at 100% fan speed. That would be acceptable performance to me.
Prime95 ONLY puts a load on the CPU, so what you are saying is entirely false.
90-100C is going to murder your CPU long-term. At 100C the CPU will be forced to underclock even more and if for some reason hits 105C the entire machine will shut down to protect the CPU.
No. Even at MAX fan speed, the CPU isn't still cooled properly enough to even hit base clock as proven by Aea's screenshots (and the various others reporting the same).
This thread has alot of "you're holding it wrong" defense...
Well, maybe actually, depends on the thermals the CPU can physically sustain. If it can sustain 85C under continuous load then that's roughly what we're seeing, not a huge amount of thermal headroom. IF it can sustain 95C under continuous load then no.
I don't really pay any attention to laptop thermals besides knowing that they run hot as f-bomb. I know that 80C is generally considered the target max temperature you want to be running at continuously on a desktop.
Does it really though? I mean it's sold as a 2.3 Ghz CPU and it only seems to briefly dip below this (indicated by the line in this screenshot)
They did however not run it for very long. We'll need some further testing.
That is Apple giving users the option to rant.
Do not worry just yet. This is only software optimisation that needs to be done, that is all.
That is Apple giving users the option to rant.
Do not worry just yet. This is only software optimisation that needs to be done, that is all.
I assume you haven't read the thread at all considering this comment is on page 1. By now it is largely established that the i9 thermal throttles so much that it actually performs worse than the i7 by semi large margin.
Thanks for sharing this, I've been waiting with bated breath for some in depth reviews of the 13" (and 15") that actually look at performance under sustained load.
I'm not proficient in German, but, it's more or less decipherable (and google translate does an excellent job albeit mangling the graphs) and the key graph (Cinebench multicore over time) can be found at about the 4/5 mark.
Looking at the results I see three key takeaways.
1. Yes, the machine throttles under heavy/sustained multicore load, but even throttled it is still 50% faster than a 2017 13" MBP.
2. Given the chassis constraints, and what we know about Intel 8th gen, this behavior isn't unexpected. It's not a defect (on Apple's part) its just a fact of life that when you stuff more cores onto the same die without significantly redesigning the architecture (Intel's knee-jerk response to AMD), something's gotta give, and in this case that something is sustained multicore performance (vs a desktop where TDP/heat don't really matter, or a design that was meant to operate with four cores in a notebook).
3. The real disappointment here is just how much slower it is than the quad core 13" Windows laptops in the comparison (~15%).
So, overall, I'd say things aren't looking bad at all for the BASE 2018 13" MBP. What I want to see now is some more in depth testing and how the i7 performs. If the i7 isn't binned significantly better, that upgrade may be significantly less worth it than it initially appears.
Now what were seeing with the i9... well, the sooner a reputable source can get out there and do an in depth review/comparison with the 15" i7/i9 looking at performance under load, the better.
[doublepost=1531893565][/doublepost]
Tell that to owners of the 2014 (and to a less degree the followup) Retina iMac(s), those things will hit 105C under sustained load... (Although they do seem to be surprisingly failure resistant at this point.)
Last edited:
At the moment, I would like to think that this is a new processor with software not specifically optimised to use it.
This does not mean I did not read the thread. Written words do not always equal truth, truth is something you find for yourself and believe in. I would be naive if I were to just assume these things so soon after a release.