Mbp normal variance in temperatures?

Question

My Macbook pro (core duo 1.86ghz) has always been extremely hot to touch, but I never thought twice about it. About a week ago I installed coreduotemp to monitor the temperatures and noticed that it reaches 96c when the processor is under load (no 3d graphics, laptop on a wooden table in a cool air conditioned room). Some parts of the notebooks cover get so hot that when a workmate accidentally touched it he screamed.
Since we have a lot of Macbook pro's where I work, I took an identical (apart from a 2.16ghz processor) model and did a profile transfer to it. When both notebooks where under load the difference in temperature was was over over 30c (64c vs 96c)! There is also a huge difference in how hot the cover of the notebook gets. The new one gets hot too, but never so hot that you can't touch it.
I called our local retailers support service, and they said they have no guidelines regarding the issue, and can not take it in for repairs.
I found this pretty surprising, so I downloaded a 3d game (sourbraten or something like that) and tried running it so that there was some processor load on the background. The screen came full of black distortion within a few minutes.
I called Apples support, and they said there must be something wrong with the video card, and the laptop should be sent for repair...
As an engineer working with servers for a living, I find this pretty funny. The support person didn't find anything strange in the fact that one of the notebooks is running at about 80% higher temperature over ambient, he actually said that "temperatures around 100c are fully normal for a notebook". I understand quality control will have some margin in a matter like this, but the difference is just huge!
The main reason I'm still wondering about this despite my mbp being accepted for repair is that we have almost 100 Apple notebooks for our workers, and there are a lot of other employees complaining about the heat. It seems most of the ones complaining received their notebooks in the same batch as me. Do we have to try to get corruption on all of them before extreme temperatures are regarded to be non-normal?

Answer 1

Hi, and thanks for all the answers!
A funny thing happened before I left work. We received a box full of silly gifts from Intel, including a usb coffeecup heater. Well the heater really works, it keeps your coffee hot until you drink it (mine was really hot for over 20 minutes)!
What's even funnier is that the notebook under load was about as hot as the coffeecup heater...
The notebook, and the one I compared it to where both updated with the latest firmware, and all automatically available operating system updates.
I'm an IT engineer, mainly maintaining fairly high end servers. I have however toyed and built computers for well over 10 years, and have quite a bit of experience with them. For example having two fully identical computers next to each other with the same configuration and running 100% load on each should result in almost the same processor temperature. In this case the cooler computer had a faster processor which should actually mean it runs slightly hotter (though speed actually does not have a huge impact on processor temperature comparing to voltage).
So what could cause the differance in temperature? There are really only a few possibilities:
- The hotter ones processor does not get good contact with the cooler
- The hotter ones fan(s) is not running as fast to remove air from the case
- A fan is broken in the hotter notebook
There are small variances in heat output between identical processors, but the differance would only be a few degrees.
I can hear the fans in both computers when they are under load. The hotter ones fan is actually running at a clearly faster speed then the cooler ones. The only logical explanation is that the heatsink does not have good contact with the processor. Due to this it can't really dissipate the heat however fast it spins the fan, and the processor remains hot. This does not explain why the case is so much hotter too... :O
I read the link you have regarding Apples guidance to temperature, but they are not actually saying ANYTHING in that article. Actually they say more in the detailed specs page you can get on the notebook on apple.com (sales). There they mention what the ambient temperature for operation can be.
Now when I called Apples support, I was told it is fully normal for a macbook pro's processor to operate at 100c. However according to Intels specifications 100c is the maximum temperature the processor is allowed to reach under any condition (this is the highest max temp on any processor I have heard of by the way). So what apple is basically saying, is that it's perfectly normal for my processor to run hotter then Intels highest allowed temperature? To add to this my ambient temperature while testing was 21c (14c under the maximum temperature allowed for operation). So what would happen if the ambient temperature went up to 35 (which is not even an especially high temperature...)? The processors temperature would most probably raise to over 110C!
Just like Intel has a maximum operating temperature for it's processors, hardisk manufacturers have a maximum temperature for their hardisks operation. Having a case too hot to touch does not make me feel very confident that the hardisk is operating under a temperature within the manufacturers limits.
But still Apple believes this is ok untill your notebook get's so hot that everything on the display becomes corrupted! Also their page concerning the macbook pro's operating temperature does nothing else then try to convince you that it's normal that your notebook burns you!
Really makes me happy I got applecare
edit:
btw, thanks for the info on the hardware test. Unfortunately I already packed the notebook for repairs, so I can't try it. Will defiantly remember it for next time though!