16" MBP: i7 or i9?

[gxxr]

Tried exporting my 167 RAWs to JPEG.
The i7 did it in 4 min 58 sec
The i9 did it in 4 min 30 sec

I was watching the fans again: The i9 gets warmer faster, so the fans ramp up to their max faster as well. The i7 stayed below 5000 rpm up until 3 minutes and a half, and only caught up to the i9's 5600 rpm towards the end.
So for short, intensive tasks, the i7 may indeed be cooler. For long tasks, like video rendering, I think they will make about the same noise - but the i9 will make that noise for a shorter time.

Btw.: I also noticed that today, the i9 is running hotter than the i7 when idle, without any apparent reason. Like the i7 did during its first couple of days. It seems that there is some setup stuff going on that I can't detect. Probably nothing major, just some little background worker tasks bumping up the temperature to around 50°C. (i7 running at ~31°C idle today)

 

[am2am]

Over

Seems to me it could be some software or hardware bug with my unit after all. Probably will try with another config, I was thinking about i9 2,3. Have you seen any benchmarks comparing it to a 2,4 in terms of temps and fan speed? My guess is 2,3 is running lower frequency so it should be cooler, but on the other hand 2,4 are higher binned, so tough to tell without actual data.

There is another thread on benchmarks where you have results on both 2.3 and 2.4
Personally I do not believe there will be huge difference between those two - unless you get lower grade 2.3

I had i9 2.3 15MBP mid 2019 and I was not very happy - benchmark performance was not on pair with others reporting here - it was bothering me. I decided to return after one week (I realised I need bigger SSD as well so it triggered my decision). I was about to order new one when the rumors of 16" appeared - so I decided to wait. Really glad I did it - it is much better machine than 15".
This time I decided to go for 2.4 to not regret after and to get higher chance of better CPU (famous binning).
Not sure if it's 2.4 or simply 16" is doing it but I have quiet machine with reasonable temps and huge performance.

 

[gxxr]

http://tbswitcher.rugarciap.com/

Would also be interesting to see how the i7/i9 compare performance wise while turbo is switched off.

I wanted to do that as the last test today, but the app doesn't seem to be working. I can click on "Disable turbo boost", but that doesn't do anything...

 

[gxxr]

Can you compare the i7 and i9 temps on idle or low/moderate workload?

Idle they seem to be exactly the same. 31°C, 5-10W. Can't really see a difference.
Moderate workloads are hard to test.

My i9 seems to have a brighter screen, so I'm 90% sure I'm sending back the i7 tomorrow. The i7 is a fine machine, and it's not that far behind the i9, but well...better is better.

 

[DRuser]

I wanted to do that as the last test today, but the app doesn't seem to be working. I can click on "Disable turbo boost", but that doesn't do anything...

Thanks for trying. You probably need allow this app to act as kernel extension. With the download comes a readme file which explains how to do it.

 

[gxxr]

Hm...I did everything like the read-me says, but I didn‘t reboot. Might still try that, but need to go to bed now. Calibrating the i9‘s screen, just to see what the Spyder thinks of it. When it‘s done, I think I‘ll call it a day. The wife isn‘t happy that I spent 2 consecutive nights in my home office
[automerge]1575494224[/automerge]
P.S.: I don‘t even have a „Allow“ option as per the readMe. Maybe the latest Catalina update broke the tool?
[automerge]1575494805[/automerge]
I just rebooted after calibrating, and still no success: I get asked to allow the kernel extension, but then I don't have any option to do so in the security panel. Smells like a bug...

 

[DRuser]

Hm...I did everything like the read-me says, but I didn‘t reboot. Might still try that, but need to go to bed now. Calibrating the i9‘s screen, just to see what the Spyder thinks of it. When it‘s done, I think I‘ll call it a day. The wife isn‘t happy that I spent 2 consecutive nights in my home office
[automerge]1575494224[/automerge]
P.S.: I don‘t even have a „Allow“ option as per the readMe. Maybe the latest Catalina update broke the tool?
[automerge]1575494805[/automerge]
I just rebooted after calibrating, and still no success: I get asked to allow the kernel extension, but then I don't have any option to do so in the security panel. Smells like a bug...

I just figured there is a newer version available that might help with Catalina (I'm running on Mojave). You also need to put the app into the applications folder. Just double clicking gave me the same effect as you described.
[automerge]1575495492[/automerge]
sorry, forgot the link: https://www.macupdate.com/app/mac/57382/turbo-boost-switcher

 

[am2am]

Calibrating the i9‘s screen, just to see what the Spyder thinks of it.

Please share your post calibration observations - was the screen far off? or contrary - out of the box was well calibrated?
This is the only area that concerns me a little - my screen is definitively warmer than my iPad Pro which I consider as a reference. Not bothering me much except I'd prefer to see my photos the same way on both..

 

[gxxr]

Please share your post calibration observations - was the screen far off? or contrary - out of the box was well calibrated?

It did get corrected quite a bit. More than my 13" at the time.
The spyder utility shows you a comparison at the end of calibration, and it was a noticable correction. It's completely neutral now. It was a bit on the warm side before (though not excessively so).
[automerge]1575533742[/automerge]

I just figured there is a newer version available that might help with Catalina (I'm running on Mojave).

Sorry, as my last step yesterday I reformatted the i7 and got it ready for shipping it back to Apple.
I felt a full week of not being productive was already too much, the i9 is being prepared for actual work now.

 

[MrGunnyPT]

Yesterday I got a bit surprised by how the i7 warmed up.

I was running a Windows VM at work most of the day due to to some SAP customisation and I noticed that the Mac was around 58-75 whole day... Keyboard started to heating up obviously and that was one of the things that I was really getting mad at.

I'm not sure if this is because I was connected to a 1080p external display and working in a VM? All I know was that I was indeed using the dGPU.

Strange behaviour, maybe room temperature being at 25c didn't help

 

[filmak]

Yesterday I got a bit surprised by how the i7 warmed up.

I was running a Windows VM at work most of the day due to to some SAP customisation and I noticed that the Mac was around 58-75 whole day... Keyboard started to heating up obviously and that was one of the things that I was really getting mad at.

I'm not sure if this is because I was connected to a 1080p external display and working in a VM? All I know was that I was indeed using the dGPU.

Strange behaviour, maybe room temperature being at 25c didn't help

Surely the external monitor and the use of the dGPU have their share.

 

[Ifti]

Im looking to go with the i9, simply because I have this problem of over-spec'ing any machine I purchase to make it 'last as long as possible' when I really don't need to. But if I get the i7 I know I wont be satisfied.....

 

[souko]

Could anyone with i7 test Cinebench R20 with turbo boost switched off? And give here some info about fan rpm and temperature, please?

 

[littlepud]

I just ran a H.264 to H.265 transcode on my new 2.3 i9... if the dGPU is inactive, I'm getting 3.4 GHz all-core sustained turbo. If the dGPU is active the all-core turbo drops down to 3.2 GHz. CPU power draw is approximately 60W and package temperature is sitting at a constant 90-95 degrees C.

 

[vebruce123]

It did get corrected quite a bit. More than my 13" at the time.
The spyder utility shows you a comparison at the end of calibration, and it was a noticable correction. It's completely neutral now. It was a bit on the warm side before (though not excessively so).
[automerge]1575533742[/automerge]


Sorry, as my last step yesterday I reformatted the i7 and got it ready for shipping it back to Apple.
I felt a full week of not being productive was already too much, the i9 is being prepared for actual work now.

so u've made your final decision to keep the i9
version instead of i7 one ?

 

[gxxr]

so u've made your final decision to keep the i9
version instead of i7 one ?

Yes I have. It was faster for tasks such as importing/exporting large amounts of RAW images, which is something I frequently do. It didn‘t seem to have any downsides (battery life etc.) either, although I didn‘t test this over extended periods of time. It is hard to do these kinds of comparisons when you only have a couple of days, and only the evening hours at that. So far, I‘m happy.

 

[Corncab44]

Yes I have. It was faster for tasks such as importing/exporting large amounts of RAW images, which is something I frequently do. It didn‘t seem to have any downsides (battery life etc.) either, although I didn‘t test this over extended periods of time. It is hard to do these kinds of comparisons when you only have a couple of days, and only the evening hours at that. So far, I‘m happy.

If I may restart this, as I'm currently making the same decision (i7 vs i9). You mentioned the i9 got hotter than the i7, and fans ramped up quicker. Is that something you till notice with extended use. I currently have both an i7 and i9 model and only noticed loud fan noise on the i9 when I was on a long zoom call. I installed Turbo Boost Switcher but seems a bit silly to have to run that if I paid for a better processor.

 

[gnasher729]

When you look at clock speeds, you are easily deceived.

Let's say you have a 2 core, 4 core, 6 core and 8 core chip with identical technology. If two cores are running, they run at the same high clock speed. For the dual core Mac, that's the speed that Apple report. If four cores are running, you get more heat and the clock speed has to go down a little bit. That's the speed reported for the quad core Mac. It's a bit lower than dual core reported speed, but it runs dual core jobs just as fast as the dual core Mac, and of course the dual core cannot run quad core jobs.

The same thing happens with six and eight cores. Every time the reported clock speed goes down, because Apple reports the speed that the Mac can run at 24/7 with all cores used. The eight core numbers are lower, but if only six cores are busy, it runs at exactly the same higher clock speed as the six core Mac.

Other manufacturers report the single core clock speed (running a single core 24/7) or even the maximum speed that a single core can run for a minute, which is even higher until it gets too hot.

 

[Corncab44]

When you look at clock speeds, you are easily deceived.

Let's say you have a 2 core, 4 core, 6 core and 8 core chip with identical technology. If two cores are running, they run at the same high clock speed. For the dual core Mac, that's the speed that Apple report. If four cores are running, you get more heat and the clock speed has to go down a little bit. That's the speed reported for the quad core Mac. It's a bit lower than dual core reported speed, but it runs dual core jobs just as fast as the dual core Mac, and of course the dual core cannot run quad core jobs.

The same thing happens with six and eight cores. Every time the reported clock speed goes down, because Apple reports the speed that the Mac can run at 24/7 with all cores used. The eight core numbers are lower, but if only six cores are busy, it runs at exactly the same higher clock speed as the six core Mac.

Other manufacturers report the single core clock speed (running a single core 24/7) or even the maximum speed that a single core can run for a minute, which is even higher until it gets too hot.

Thanks. So the reason an 8 core would run a fair bit hotter than the 6 core i7 is the clock speed, yes? Even on non-intensive tasks my i9 seems to run a lot hotter than the i7. Which is curious if shouldn't need to run at such high clock speeds?

I know that battery life is actually a bit better on the i9 when pushing to the limits, but what about everyday use? If it's running hotter then it must be using more battery?