Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
can you post a screen shot from iStat ?

Here you are:

capturedcran20110608000.png


After 30minutes of video and audio encoding:

capturedcran20110527183.png
 
For people who have an external display, don't expect temps as low as what you're seeing here as the dedicated GPU is always turned on generating some heat.
 
Here you are:

i refuse to accept that idle (finder only, no external display) 47-50C of my system 2.3GHz is acceptable seeing your temps.

i'll give it another week and if AS5 won't get any significant improvement (i'ts been about 200 hours ON since application) will try another attempt with OCZ Freeze.

comment on spreading vs. grain of rice spread with the heat sink ? what did you do ? what's better method ?
 
Even though it runs at the same temp as many many many other identical computers....

that's simply not true. as the lucky above get 36C idle with same system.
i love this comp, best MBP yet, but it bothers me that i cannot get the temps that others report.
 
the lucky above get 36C idle with same system.

Right. The key word here is "lucky". The vast majority (me included, and I have a 2GHz processor and standard def screen) have temps very similar to yours.

And do you notice that guy has a weather widget on his toolbar. 15 degrees. Is it that cool where you live?
 
Right. The key word here is "lucky". The vast majority (me included, and I have a 2GHz processor and standard def screen) have temps very similar to yours.

And do you notice that guy has a weather widget on his toolbar. 15 degrees. Is it that cool where you live?

I was just about to remark on this. Of course! He's clearly somewhere in France, it's colder there and temps will be lower.
 
Right. The key word here is "lucky". The vast majority (me included, and I have a 2GHz processor and standard def screen) have temps very similar to yours.

And we all have a year's warranty where Apple will fix or replace the MBP should it overheat and stop working. At best, you might end up with a machine that's a little quieter under full load, but when idle you won't hear the fans whether it's running 30 degrees or 50
 
I was just about to remark on this. Of course! He's clearly somewhere in France, it's colder there and temps will be lower.

surely, just because there is 45C temps outside, you don't sit in a sauna in your house do you ?
same the other way. that's why heating/cooling exists.

* my head on a plate if i'm wrong a he sits in 15C room at his MBP
 
surely, just because there is 45C temps outside, you don't sit in a sauna in your house do you ?
same the other way. that's why heating/cooling exists.

* my head on a plate if i'm wrong a he sits in 15C room at his MBP

I'm assuming most people try and keep their homes at around 21-22°C. However, I know from experience in the UK (no air-con) that temperatures indoors can hit up to 30°C during a hot day. At most though, I reckon it would make about 5° difference, probably more like 2-3°.
 
i refuse to accept that idle (finder only, no external display) 47-50C of my system 2.3GHz is acceptable seeing your temps.

i'll give it another week and if AS5 won't get any significant improvement (i'ts been about 200 hours ON since application) will try another attempt with OCZ Freeze.

comment on spreading vs. grain of rice spread with the heat sink ? what did you do ? what's better method ?

I don't do anything up till now. My computer is stock (2,2GHz, 128SSD and Hi-res)

@clickclickw00t: yes I'm in France; it's beginning of summer but weather was not cold at all these last weeks. Nevermind the temp inside the room was around 21° when the screenshot was taken (in the middle of a rainy night, this explains the 15° outside given by the toolbar widget). When it is 25° in the room it is sure that result is not the same....
 
Last edited:
I don't do anything up till now. My computer is stock (2,2GHz, 128SSD and Hi-res)

@clickclickw00t: yes I'm in France; it's beginning of summer but weather was not cold at all these last weeks. Nevermind the temp inside the room was around 21° when the screenshot was taken (in the middle of a rainy night, this explains the 15° outside given by the toolbar widget). When it is 25° in the room it is sure that result is not the same....

this morning was 19C outside. around 22C room. idling still around 46C.

my point is: did i not apply the paste correctly ? or am i just unlucky that the chips are "bad batch" and won't run as cool as others ?
 
I had a long post all typed up but after re-reading it, I thought it came off as a bit patronizing. I don't want to patronize you, because I don't mean to upset you. In my opinion, you're just looking at this from the wrong angle. Most users' CPUs (mine included) run at very close to the same temp as yours. That doesn't mean they're "bad". We're not unlucky to have CPUs that idle ~45 degrees or so. Rather, I'd say that Tom Lopez is "lucky" that his runs that cool.

Now, doesn't make any difference in performance or reliability or longevity or anything related to that kind of thing? No. A computer that idles at ~45 will perform exactly the same as an identical computer that idles 10 degrees cooler.

On the other hand, take my old 2008 aluminum Macbook. It had been idling at ~85 degrees. That is hotter than it should be. So I re-did the thermal paste on it, and the idle temps are now down to ~58 degrees. That's a reasonable temp for that model macbook.

The fact that the temps dropped showed that the original thermal paste wasn't doing it's job very well. The fact that your temps are roughly the same before and after thermal paste reapplication shows that the original thermal paste in your computer were doing there job satisfactorily.
 
Last edited:
IC diamond

So I've replaced my TP a week ago with IC Diamond, nothing spectacular, but still very noticeable results, idling at 42C with 22C outside (used to be 50-55), surfing right now with 48-52C (about 5 tabs open, used to be 60-65), I also run Kaspersky antivirus (don't judge, I'm a former PC user and am paranoid about viruses etc...) so that eats up a bit of resources. Playing SC2 I run at about 77-84C (used to be 90-95), am on 2.3GHz cpu. Overall am happy with the results, it didn't take effect right away, am guessing because of how thick IC diamond is, and when it heats up it spreads more, so I guess every time the cpu heated up it spread more and more until it hit it's thinnest point. I applied a pea sized amount on GPU and a small strip on CPU and then squished it. I am actually thinking of machining a pure copper heat sink LOL any ideas about that, worth it?
 
i've no luck getting any better results. yesterday i tried to apply paste 3 times! :mad:

first with Shin-Etzu G751 Thermal paste and then with OCZ Freeze (pic below)

seems there are 2 camps here:

1) those who recommend to spread the paste in thin layer
2) those who recommend to squeeze a line/drop and let it spread with the heat-sink (should produce less air bubbles ?)

i tried the Shi-Etzu as spread, then let the line spread with the heat-sink.

the last attempt was with OCZ Freeze - see pics that i actually tried to spread thin with a plastic piece of a business card.

the temps are very disappointing. the ambient in my room is about 24C

this is idle on integrated GPU:
Image

how can someone get 31-33 CPU as posted earlier ?

what am i doing wrong ? any pointers ?

I have had computers for 20 years now, and for 15 years I've replaced the thermal paste, used to do it the 'spread' method, but have recently converted into a pea sized squish method, as that apparently spreads it super evenly (if proper amount is applied) without any air bubbles - see when you spread the paste (unless you can spread it evenly, to the micron, on the entire surface) when the heatsink touches down on the paste, air gets trapped, air is not a good heat conductor, enough of it, is actually worse than what chinese are doing with overapplying the paste.
 
Correction!

So I've replaced my TP a week ago with IC Diamond, nothing spectacular, but still very noticeable results, idling at 42C with 22C outside (used to be 50-55), surfing right now with 48-52C (about 5 tabs open, used to be 60-65), I also run Kaspersky antivirus (don't judge, I'm a former PC user and am paranoid about viruses etc...) so that eats up a bit of resources. Playing SC2 I run at about 77-84C (used to be 90-95), am on 2.3GHz cpu. Overall am happy with the results, it didn't take effect right away, am guessing because of how thick IC diamond is, and when it heats up it spreads more, so I guess every time the cpu heated up it spread more and more until it hit it's thinnest point. I applied a pea sized amount on GPU and a small strip on CPU and then squished it. I am actually thinking of machining a pure copper heat sink LOL any ideas about that, worth it?

Actually, huge correction, when I looked at the idle temps of 42C before, that was with uTorrent running (downloading 10 different things at over 800kb/s), I killed it, left my 15" MBP idle for about 20 mins, came back and the cpu was running at a nice and cool 33 Celsius! :D
 
I have had computers for 20 years now, and for 15 years I've replaced the thermal paste, used to do it the 'spread' method, but have recently converted into a pea sized squish method, as that apparently spreads it super evenly (if proper amount is applied) without any air bubbles - see when you spread the paste (unless you can spread it evenly, to the micron, on the entire surface) when the heatsink touches down on the paste, air gets trapped, air is not a good heat conductor, enough of it, is actually worse than what chinese are doing with overapplying the paste.

makes sense. last application was spread. will try again soon with the grain of rice method.
 
Actually, huge correction, when I looked at the idle temps of 42C before, that was with uTorrent running (downloading 10 different things at over 800kb/s), I killed it, left my 15" MBP idle for about 20 mins, came back and the cpu was running at a nice and cool 33 Celsius! :D

did you do manual spread or the rice of grain spread with heatsink ?
 
OK, I took the plunge and reapplied my thermal paste. I used Arctic Silver 5. I was going to use Artic Silver Ceramique until I took the MBP apart and saw the stuff on there wasn't white and didn't want it to look too different from what was on there already (aside from the fact there is now considerably less compound than before I started and it's a different shade of grey ;)).

Worst moment was with the small RF shield by the power connecter, where one of the screws didn't want to come out and I didn't quite have a screw driver small enough to fit it. I managed to rotate the shield around a bit, which loosened the screw enough to get it out with the screw driver I did have.

My idle temperature now is running at 33-36°C, room temperature is 21.6°C. I ran a Prime95 load, which still got it up to 91°C, although the fans topped out at about 5800RPM, instead of the 6200RPM they usually reached.

Definitely an improvement, I just hope I don't need any warranty repairs on my logic board, or things could get a bit sticky.
 
Because of turbo boost, load temps will always get to the 90-91C range because the CPU is "overclocking" itself for increased performance. In theory, with better thermal paste, you should stay overclocked higher for a longer period of time from a base of 2.0-2.2 to upwards of 3.4 Ghz (depending on the number of cores active). Idle temps will get lower over time, though I wouldn't expect much. The greatest benefit to replacing your thermal paste is lower fan speeds.

And like I've said before, if the dedicated GPU is turned on, base temp is going to be around 60-65C regardless.
 
Having redone the thermal paste (and got a significant improvement), is there any point in taking out the Applecare extended warranty on my MBP? I realise there are those who say that it voids the warranty and others who say your warranty is only void if you break something.

This section seems to be relevant on two counts:

Apple said:
This warranty does not apply: (a) to consumable parts, such as batteries, unless damage has occurred due to a defect in materials or workmanship; (b) to cosmetic damage, including but not limited to scratches, dents and broken plastic on ports; (c) to damage caused by use with non-Apple products; (d) to damage caused by accident, abuse, misuse, liquid contact, fire, earthquake or other external causes; (e) to damage caused by operating the product outside the permitted or intended uses described by Apple; (f) to damage caused by service (including upgrades and expansions) performed by anyone who is not a representative of Apple or an Apple Authorized Service Provider (“AASP”); (g) to a product or part that has been modified to alter functionality or capability without the written permission of Apple;

(f) only seems relevant if applying the paste causes any damage, but (g) could also apply as well (i.e. modified capability).


Has anybody had a warranty repair on a logic board following replacing the thermal paste? Was anything said? Do replacement logic boards come with heatsinks already attached, in which case they are unlikely to remove them to check?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.