2018 i9 MacBook Pro throttling discussion [merged]

[Cervontez]

Hey guys, there is a new great video on youtube which shows that the i9 does not really throttle that much. It actually looks quite promising . Have a look

Although this test shows that there is almost no throttling, it also shows that the laptop is never capable to turbo boost (i.e. it never goes beyond the 2.9ghz base clock) ?? That's not a good indicator!

What do you guys think? Still, I cannot get my head around this other video https://www.youtube.com/watch?v=ip-sZfWaVo0&frags=pl%2Cwn in which the new i9 is slower than the top spec 15" 2017 model.

What do you guys think??

Am I right to assume this would be closer to real world usage vs doing a stress test taxing all cores? It appears the tests, other than the stress tests, showing the worst of the i9 are using software that is notorious for not running great on Macs ie Premier.

 

[lambinho]

It's ok. Regarding the video, it seems the 2018 MBP is doing really good on FCP X and DaVinci compared to Premier Pro. Despite the throttling the i9 is still performing better than not just the 2018 i7, but also the 2017 i7.

It’s performing better yes. But if you ramp up the fans and change the tdp to 45w on the i7 2,6 it’s getting much higher scores.
We have not seen any result of the i9 doing better with locking it at 52w, but I would be suprised if that is not the case. (It is doing worse at 45w)
But an update from Apple can fix this, it’s just a mather of time in my opinion.

 

[Aea]

Volta makes a huge difference. I'm going to run a more scientific series of tests and post them later but it looks like I'm seeing gains of 150 on my cinebench just be disabling turbo boost and capping the CPU to 45W.

Will update.

 

[M.Rizk]

Volta makes a huge difference. I'm going to run a more scientific series of tests and post them later but it looks like I'm seeing gains of 150 on my cinebench just be disabling turbo boost and capping the CPU to 45W.

Will update.

If you have Bootcamp please give Intel XTU a try too and set it to 52W (if you have an i9). Again, I would like to thank you personally for the videos and tests you've done so far

 

[Aea]

Will try. Got any recommended video recording software for windows?

 

[M.Rizk]

Will try. Got any recommended video recording software for windows?

This seems to be the easiest/free way. http://www.itprotoday.com/windows-10/how-use-windows-10s-screen-recording-utility

 

[linkgx1]

I'm starting to get into web design/ development. I'm getting the base 2018. Would the throttling affect me?

 

[lambinho]

I'm starting to get into web design/ development. I'm getting the base 2018. Would the throttling affect me?

The 2,2 don’t seem to have the same issues, so I would say no.
It’s not close to the throttle level of the i7 2,6 and i9 2,9.

 

[bigcat]

Can someone summarize the changes needed for a 2.9 i9 to prevent the catastrophic throttling and what apps to use (I’m talking about limiting voltage, fan settings, etc). Maybe add this info to the first post?

 

[content]

Is this what you are asking for?
#946

The Notebookcheck review was done before they themselves discovered the power limit fix.

I am not sure, but it appears they are acknowledging the 13% gap between the mid model 6-core 2018, and the base model 4-core 2017. And by limiting power consumption to the processor on the 2018, this gap widens. The 2018 now sees 20% more performance. However, at the cost of closing the gap between the two when doing every day work that does not tax the processor. Further it appears, they would need to limit power consumption to the 2017 as well, to make the test completely fair. I mean, the gap may get close to 13% all over again, if they did this. If they applied the same tweak to the 2017 as well. I am not sure, but it still does not seem conclusive?

First article:

https://www.notebookcheck.net/The-n...nd-clearly-beaten-by-the-XPS-15.317264.0.html

Second article:

https://www.notebookcheck.net/Apple...e-performance-with-a-few-clicks.317552.0.html

 

[M.Rizk]

I am not sure, but it appears they are acknowledging the 13% gap between the mid model 6-core 2018, and the base model 4-core 2017. And by limiting power consumption to the processor on the 2018, this gap widens. The 2018 now sees 20% more performance. However, at the cost of closing the gap between the two when doing every day work that does not tax the processor. Further it appears, they would need to limit power consumption to the 2017 as well, to make the test completely fair. I mean, the gap may get close to 13% all over again, if they did this. If they applied the same tweak to the 2017 as well. I am not sure, but it still does not seem conclusive?

First article:

https://www.notebookcheck.net/The-n...nd-clearly-beaten-by-the-XPS-15.317264.0.html

Second article:

https://www.notebookcheck.net/Apple...e-performance-with-a-few-clicks.317552.0.html

As I mentioned earlier, English is not a first language for me but what I understood from the 'few clicks" article is that Apple in fact used to enforce a power limit pre-2018 and not just depend on the thermal throttling. If this is true then the 2017 CPU is already fully optimized by Apple.

45W is probably not even the sweetest spot for the i7, or the i9 but that is unfortunately the highest value available on the Volta app in the mean time.

This is all about the raw performance. Overall performance has probably a much bigger gap due to the 2018 having DDR4 RAM and better properties in the 8th Gen CPUs in general.

 

[M.Rizk]

2.2 GHz i7 15" and Base 13" results after applying power limit by @dof250
#30

 

[Howard2k]

This scandal has blown up WAY out of proportion. How difficult it to download a fan speed app and turn the fans up to 100%

If you are spending this amount of money on an MBP i9 and you don't know how to download a fan speed app, then you shouldn't be buying this computer.

Agreed. Expecting that Apple would have figured this out is just absurd. This is of course something that consumers should be taking on.

 

[Aea]

Fan curves are not the problem. Will back up with data in a few hours, doing some fairly rigorous testing and it's taking a while.

 

[fatalogic]

I'm starting to get into web design/ development. I'm getting the base 2018. Would the throttling affect me?

You can do web dev on almost anything with a screen and a keyboard. A MacBook Pro will be like taking a Ferrari around a city block when it comes to beginner web dev.

 

[Elektrofone]

Fan curves are not the problem. Will back up with data in a few hours, doing some fairly rigorous testing and it's taking a while.

Thanks. Which devices are you comparing?

 

[M.Rizk]

Thanks. Which devices are you comparing?

i9 according to his YouTube channel: https://www.youtube.com/channel/UCE4_1i3A1EpvwcWMCHUztJQ

 

[Ploki]

This scandal has blown up WAY out of proportion. How difficult it to download a fan speed app and turn the fans up to 100%

If you are spending this amount of money on an MBP i9 and you don't know how to download a fan speed app, then you shouldn't be buying this computer.

I should also assemble it myself i guess.

 

[Elektrofone]

i9 according to his YouTube channel: https://www.youtube.com/channel/UCE4_1i3A1EpvwcWMCHUztJQ

Core i9 and the i7?

 

[M.Rizk]

Core i9 and the i7?

2018 i9 and 2014 i7.

 

[Aea]

I am testing on the i9 / 32 GB / 1TB.

 

[kylepro88]

OK this is the first time I've seen the CPU operating at these clock speeds and temps... nobody online seems to be using the Premiere enabled hardware acceleration on H264 export.

**go 6min in**

 

[phato777]

I'm starting to get into web design/ development. I'm getting the base 2018. Would the throttling affect me?

I've been a web developer for ~15 yrs now and also just got an i9. None of the new Macbook builds will hinder your development, throttling or not, even if you're compiling code. I love my new heater.

 

[M.Rizk]

OK this is the first time I've seen the CPU operating at these clock speeds and temps... nobody online seems to be using the Premiere enabled hardware acceleration on H264 export.

**go 6min in**

That is because of Intel Quick Sync. Most other YouTubers tend to turn it off, some say to test the actual CPU cores performance, and others claim it produces videos of lower quality. Idk, I know nothing about video editing.

 

[kylepro88]

That is because of Intel Quick Sync. Most other YouTubers tend to turn it off, some say to test the actual CPU cores performance, and others claim it produces videos of lower quality. Idk, I know nothing about video editing.

Right I get that. And the CPU utilization is low... but for those talking about the CPU performance under Premiere... it helps level playing field. FCPX uses QuickSync too. I've never heard anybody say it produces lower quality videos though.