2018 i9 MacBook Pro throttling discussion [merged]

[M.Rizk]

Hear hear.

I joined this forum yesterday (but been lurking for a few days) to monitor how this whole episode unfolded, and I'm super glad I did. Had I allowed myself to be swayed by doom-mongers on YouTube and Reddit I would have likely cancelled my order.

There is nothing wrong with canceling an order and keeping an eye on the scene before making a final decision. It is not like they were cheaper the other day and now more expensive.

Who knows, maybe this pushed Apple to release a fix fast. All a business cares about is money.

 

[anshuvorty]

I found this forum was the best place to go to watch all this unfold, and to get all the reliable data posted elsewhere - links to videos, knowledgeable critiques of the videos both good and bad, and all the news. This is clearly the best place to go for the real truth, and there are some really smart/experienced people here. The bad behavior was at a bare minimum, and I learned a lot too. I am grateful to be a small part of this community, I appreciate learning what was really going on.

This forum was likely a significant assist to Apple to get this thing fixed (I was hoping for a fix by September actually). I posted this because there are likely many of us lurkers who could not really contribute to the dissection of the problem, but benefitted greatly by being able to follow along.

Kudos to @winterny but many other smart contributors here too, job well done and thank you!

Well maybe I was lurking around in the wrong threads then. Some of the threads posted here got quite toxic. It was something else. But I agree that when it mattered the most, the MR community came together and did the correct job of collecting the crucial data and working on their own 3rd party fixes before Apple released the official fix.

 

[GerritB]

Just joined last week after ordering my i9 and hearing about the problems. Been a pleasant experience so far and learned a lot. I'm really looking forward to the next couple of days to see how benchmarks go after this update to see if I'll keep my i9. It really has to translate into better performance though, for now it doesn't seem that way.

 

[LordGeddon]

There is nothing wrong with canceling an order and keeping an eye on the scene before making a final decision. It is not like they were cheaper the other day and now more expensive.

Who knows, maybe this pushed Apple to release a fix fast. All a business cares about is money.

A fair point, but I was alluding more to my own fickleness with things I know very little about. I simply saw an onslaught of YouTube videos and self-proclaimed tech bloggers calling scandal and was ready to pull the plug there and then. Instead here I got a more objective breakdown of the limitations and underlying causes of the issue and found myself learning a fair bit.

I still have my two week grace period to gauge whether its been a worthwile purchase for me, and beyond that I've got AppleCare, so at this very moment I'm feeling pleased.

 

[Aea]

 

[Nozuka]

Who knows, maybe this pushed Apple to release a fix fast. All a business cares about is money.

True, but also not true.

There are real people working on this. And if i was part of that team i would definitely want to fix this as fast as possible. It would be very disappointing to me to see results like this with the device i was working on. Surely these people have some pride in their work too.

 

[Elektrofone]

Thanks for posting. Interesting results. Seems like the i9 isn't doing much better than the i7 models in Cinebench. Would love to see some real-world usages like in creative applications.

 

[Hardy1989]

.

 

[LordGeddon]

How would one go about benchmarking 'real-world' usage?

I've just got my i9 MBP here and was curious as to how I'd measure that myself. Is it purely anecdotal in a 'oh hey everything is pretty snappy and my code compiles a couple seconds quicker' sort of way? Or is there a more quantifiable way of measuring it?

 

[wallysb01]

Thanks for posting. Interesting results. Seems like the i9 isn't doing much better than the i7 models in Cinebench. Would love to see some real-world usages like in creative applications.

They look pretty similar to me. 1050-ish on the first run or two, then stabilizing around 950-1000 from there.

A user on the 2.6 model here.

A user on the 2.2 model here. Well that's just right below the first one, but there are others on that thread seeing similar things.

 

[Aea]

How would one go about benchmarking 'real-world' usage?

I've just got my i9 MBP here and was curious as to how I'd measure that myself. Is it purely anecdotal in a 'oh hey everything is pretty snappy and my code compiles a couple seconds quicker' sort of way? Or is there a more quantifiable way of measuring it?

This is hard to do, and it really depends on what your usage is.

If you have a repeatable workload then running it in a way that you would run it in real-world usage would be the most reliable way.

i.e. If you're compiling something every once in a while do a bunch of runs (giving the system time to cool in between) and looking at the compile time difference would be an possible approach.

 

[LordGeddon]

This is hard to do, and it really depends on what your usage is.

If you have a repeatable workload then running it in a way that you would run it in real-world usage would be the most reliable way.

i.e. If you're compiling something every once in a while do a bunch of runs (giving the system time to cool in between) and looking at the compile time difference would be an possible approach.

Ahh, but then I'd need to find a 2.2/2.6 model, mirror the same configs I use and run them on those devices? Ahh, lol.

 

[galactic orange]

Got my i9/32/560X/2TB today. Ran Power Gadget with my Macs Fan setup. I took this screenshot right after it finished installing.

Results during installation of 60GB of libraries into Logic Pro X, after installing/downloading 520GB of NI Samples/VSTs/Plugins with similar results.

More testing later when I start bouncing big audio files. Let me know what you think.

I still vote that its a fan management problem, coupled by the tight chassis. I took it into my own hands, granted, utilization was very low, so it is a hot potato, but with proper cooling it screamed.

EDIT: In this post I'm really just pointing out that it never went under 4GHZ during operation while I was downloading big files, transferring files from two different drives and boring the interwebs. I'm not torture testing this machine for the sake of it... just showing routine operation while playing with my fans speeds.

Thanks for the Logic Pro X info. It would be great if someone could post some realtime DAW statistics after applying Apple’s fix for any of the 2018 MBPs.

 

[rubberducker]

Thanks for posting. Interesting results. Seems like the i9 isn't doing much better than the i7 models in Cinebench. Would love to see some real-world usages like in creative applications.

take it back dude. you have been fretting for days in here. your i9 is going to be a little bit better than the i7 - but it won't be night and day. If money is a worry or its going to keep you up at night - just swap it out.

I have the i9 and was pretty sure Apple would fix it (it would be very "un-apple" like not to have some "gate" or other around a major new release). And was always going to be storm in a teacup. I use AE and Premiere all the time and am not expecting miracles - just a steady bump in performance. The i7 wont be very far behind whatever we both get on the i9 id wager.

 

[wallysb01]

Ahh, but then I'd need to find a 2.2/2.6 model, mirror the same configs I use and run them on those devices? Ahh, lol.

Yeah, this is why average users don't really do more than post results to standard benchmarks and we can pool results across users with different machines. Unless you want to truck down to the apple store, drop 10K on 3 machines, then return 2 of them? What's really most useful for you, would be to do the same common job on your new machine and your old machine and compare how much better it will be for you.

 

[Standard]

Would be great if you guys could do the Unreal light baking test again with this patch!

 

[KarmaRocket]

Would be great if you guys could do the Unreal light baking test again with this patch!

Yes, would like to see this test as well. thanks

 

[LordGeddon]

Yeah, this is why average users don't really do more than post results to standard benchmarks and we can pool results across users with different machines. Unless you want to truck down to the apple store, drop 10K on 3 machines, then return 2 of them? What's really most useful for you, would be to do the same common job on your new machine and your old machine and compare how much better it will be for you.

Trouble is my old machine is a 2011 MBP, so I should hope my new one beats it by some distance lol.

 

[aevan]

True, but also not true.

There are real people working on this. And if i was part of that team i would definitely want to fix this as fast as possible. It would be very disappointing to me to see results like this with the device i was working on. Surely these people have some pride in their work too.

So true. And something a lot of people on this forum fail to realize. Generally, people like to think everything is always about money, but even in a profit-driven world, it's not just about that.

There are people working hard and passionate on all these projects and they want to do good. These people do have pride and would quit if there was deliberate sloppiness, planned obsolescence or plans to make something intentionally suck.

It's easy to imagine any corporation as a greedy, cold entity, but these companies are made out of living, breathing humans who, especially in this case, have many alternatives where to work. These people didn't sign up to do bad work.

Of course, cynicism has become so prevalent among enthusiasts, forums and journalists, that a lot of people will just dismiss this and continue the same narrative generated by the Internet echo chamber.

Also, to those who will respond with "if there weren't people who noticed this, Apple wouldn't have fixed it" - please don't compare the conspiracy-theory rants and whiny wannabe-expert comments I'm talking about with honest, good natured criticism, like the one frome Dave2D. Of course that a critical view is welcome.

 

[dan9700]

Bootcamp still isn't fixed I hope they fix it

 

[blackberrycubed]

Guys/Gals, pull out your MBPs from the freezer, apply patch and do not put i back in freezer. Patch has bultin in freezer cooling included!

 

[Blue Sun]

I think I’ll be picking up a 2018 MBP after all

 

[AppleExperte]

I just bought the 15inch 2.6/16GB/560x/512GB model a few hours ago. After I left the Apple Store the push notification came in that there is an update that improves the performance. In that moment my mood went up and when I came home I immediately installed the update. I ran 5x Cinebench in a row, first one 1079, after that all tests were about 1030. The fans were at full speed (about 5500rpm; source: Mac Fan Control), even after the benchmark till it reached 45-50 degrees celsius (almost idle temperature).

After these tests I unplugged my MacBook and 1:20 hours later I am still at 90% (Watched Youtube, installed some light programs, tested Netflix, improved some setting). That is very impressive! In activity monitor it shows me that there are 10-15 hours left (changes every time I do something different).

So far so good. I hope that it won't change in the future (improvements are always welcomed).

Greetings from Germany!

 

[LeeW]

59 pages on it's a result that has hopefully calmed everyone back down

 

[sennomulo]

Max Yuryev has a new video testing out the bugfix on his i9: