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.
How to test MBP 2018 for possible heat, noisy fan and thermal throttling issues? Let's make tables.
- Thread starter hajime
- Start date
- Sort by reaction score
It's still a poor comparison - comparing a two year computer against a current model.
Ok, from heat point of view, i got this today
2018 all cpu:
- 2.2GHz 6-core Intel Core i7 around 40.5C under load
- 2.6GHz 6-core Intel Core i7 around 42 C under the same load
- 2.9GHz 6-core Intel Core i9 around 45 C under the same load...almost never hit that 4.8Ghz
So i chose the base cpu 15" Mbp...i dont know why Apple even put the i9 there...without a different cooling system
2018 all cpu:
- 2.2GHz 6-core Intel Core i7 around 40.5C under load
- 2.6GHz 6-core Intel Core i7 around 42 C under the same load
- 2.9GHz 6-core Intel Core i9 around 45 C under the same load...almost never hit that 4.8Ghz
So i chose the base cpu 15" Mbp...i dont know why Apple even put the i9 there...without a different cooling system
Thanks. So they are different by a few degrees. In real world, is it significant? I am not sure but 4.5c difference may make a noticeable difference.
Are you able to get numbers on how noisy they are under normal and heavy loads? I don't want to get a MBP that sounds like a vacuum cleaner or a jet engine like my MBP 2010.
Without adequate cooling and power the i9 is little more than bragging rights. I wouldn't be surprised if the base 2.2GHz 8750H proves to be the best option given my own W10 notebook easily passes 1200CB under sustained heavy load with the 8750H. Apple's i9 is in the mid to high 900CB with the CPU likely needing to roll back performance further as temperature rises...
Asus new Zenbook Pro suffers the same big numbers, inadequate cooling & poor comparative performance.
Q-6
I think your opinion of the i9 Macbook Pro will likely turn out to be accurate, its potential hampered by insufficient cooling from the chassis. I am most interested in whether the mid-tier 2.6 GHz can turbo without throttling or triggering the fans too much. Its base 2.6/4.3 GHz clock speeds aren't that much higher than the base 2.2/4.1 CPU so maybe the form factor can handle its heat.
I know some people on YouTube have i9 but they don't talk about heat/fan/thermal throttling.
TBH I'd like to be proved wrong, equally the base 2.2GHz 8750H can pull as much as 90W if configured so. Better binned yes, however the laws of physics remain.
My own Asus GL703GS with the 8750H holds Turbo strongly, that said it's a larger chassis with a very robust cooling solution, and 12V fans to maintain optimal performance of the 8750H & GTX 1070.
Q-6
I think we need to see more thorough benchmarking with sustained workloads with 1, 2 ... 6 cores active and see what frequencies and temps are maintained. I'm not at all surprised that the i9 didn't hit 4.8 GHz if the workload with 4+ cores....
I heard that for the i9, not all cores get the max GHz. Only one does. Is this really true?
That's the key sustained workloads, these rapid fire tests are pretty much useless. My 8750H can hold 3.9GHz, only the likes of Prime95 Small FFT stress test will force the CPU to reduce frequency. Thermals are fine at around 70C, more the enforced PL-1 power limit of 45W (PL-2 being 90W)
As the more detailed unbiased reviews are released the picture will be far clearer regarding the 2018 MBP performance.
Q-6
Basically yes, same as all other Intel Core CPU's. More fully loaded Core's lower the frequency.
Q-6
Actually yes it is. There are so many different ways to test computers for heat, noisy fans, and thermal throttling. The hard part is making accurate, consistent tests to see if there really is thermal throttling, and if one or more of the new Macs is more susceptible than another.
Also, keep in mind that this is something that will very across the different computer sizes and CPUs specs. Also, even once the tests come in, while the numbers will differ, each model year will have different heat allowances. It is highly likely that the 2018 models will run hotter, but that won't be surprising.
Throttling will probably be the most reliable testing, but it still may not tell you if the new MacBook Pros are a good fit for you. Often times heat and throttle tests are done by running the CPU on the highest/hardest setting it can go. With these new quad and hex core processors you are not likely to hit that max level long enough to cause throttling unless you are doing heavy 4k video editing, mass photo batch editing or VR development. Most users shouldn't experience excessive throttling.
Thanks. So the way to go is to take an educated guess on the best configurations that fit my usage. Then try it for two weeks and judge whether to keep it or get another configs and try and repeat.
Honestly that is my opinion. I would say that i7 is fine, the i9 is probably more likely to experience throttling because it is more power hungry, and the performance boost over the i7 will be negligible to most users.
I forgot to add that the other challenge of following those tests has to do with even little details like where the MacBook is placed while doing the throttling tests. I use my computers on a glass top desk, which will help dissipate heat better than a wood table. Stuff like that has to be taken into consideration.
Ultimately buy one you think fits your needs, test it for a week, and if you don't like it or think it doesn't suit your needs, return it and try something else. Sometimes reading all the review and forum threads can just make it more confusing/frustrating than just buying the one you think best suits your needs and seeing how it feels.
The cache of the i9 is considerably larger then the i7, I'd imagine the i9 will seem enormously faster for every day things, even at single core. I might be biased due to having ordered the i9 blindly, but I really think the cache is the performance gain dark horse that no one is considering right now.
This is very interesting... anyone with experience regarding this topic ? Im also deciding wether I get the 2.2, 2.6 or 2.9.
This is only because FCP hasn't been optimized for for 8-gen chips in the Mac. Same thing happened with the 2014 and 2016 model MacBook Pros.
Also an example on how much of a difference my i9 MacBook Pro made for me is when applying multiple filters in my workflow compared with the workflow of a friend of mine who just got his i7 MacBook Pro was a whole minute and a half better on my computer. Now multiply that with the amount of work I do every day, I would save hours in a year.
Also, think about this, Apple would not put the i9 there for a stupid reason just for the sake of it and have the i7 beat it. This would in this case be another action class lawsuit. the i9 does make a difference. Let me see if I can do a review with my friends and mine and do a proper comparison.
Last edited: