Unfortunately I'd point at ASUS this time, considering that literally each and every ASUS notebook that family/friends/colleagues have owned have all have inconsistent hardware related faliures. Not a single one is happy with their ASUS notebook after a few months as-is.
(the one exception has a caveat, the guy re-assembled his ultrabook, and is the only happy one left)
I just don't think anything Pegatron builds is of a suitable quality for someone that places any demand on their system.
Condolences with your woes Michael :-(
I suspect that you may have to check that everything is connected correctly, and ensure that the thermal paste is applied correctly (hence, re-assemble your machine)... Not ideal :-(
Broadwell would solve thermal issues.
Michael, wonder if it's not a power issue. My exact same model zenbook has given me a TON of issues.
But it seems mostly fixed now after 5+ motherboard replacements and one battery replacement. Think it was the battery replacement.
Would randomly reboot, randomly power off, and go into deep sleeps where it can't wake up again, like at all.
Anyway, now it's stable after last repair (except keyboard backlight no longer works and mic, joy).
I monitor my CPU in KDE bar and my temp is normally 50 degrees C most of the time with a ton of things open. It can burst to 70-90+ degrees but doesn't cause any issues.
I have installed and enabled thermald on Arch and I believe that has helped keep the temp down.
I seriously think Unity is the reason. I've seen A Lot of laptops overheat because of Unity, if they switched to anything else it was working all good...
Unity lol. If you don't install drivers OpenGL is run on the CPU.
Run a memory check and test an alternative DE, both are quick and non-intrusive tests. If random reboots still happen, I would RMA it.
Actually last I checked thermal issues are only getting worse as Intel goes down manufacturing nodes. I think the density might just be "concentrating" the heat to one spot more, increasing the difficulty of cooling it.
Originally Posted by GT220
Originally Posted by rikkinho
I agree. After switching out to MX-4 my laptops' (yes, two of them!) have seen on average 10*C reductions in temperature. The original thermal paste on them was like cement. In fact I think actual cement could have been better than the junk they put on.
Originally Posted by leonmaxx
FWIW, I did upload pictures of the process if anybody is interested:
I even replaced the thermal paste on my Chromecast too, and it dissipates heat well enough where the outside of the case doesn't feel boiling hot anymore.
With all do respect sir, you have applied the thermal past wrongfully. Your method would produce bubbles of air between heatsink and the Processors. You have to put a large grain of thermal paste in the middle of cpu and let the heatsink rub it on the cpu.
Originally Posted by MWisBest
That may be true for CPUs with an IHS, but these are bare dies, so just counting on it to spread itself out is not the way to go here. The line between too little thermal paste and too much thermal paste is just too small.
Originally Posted by keivan
The air bubbles can be accounted for by how you mount on the heatsink, in this case I didn't simply put it straight on there but instead slowly made contact via a diagonal (e.x. instead of just || it was |\ and then ||) so as to let the air get out as I went along. I could clearly see the air escaping as I did so. I do agree that it's not perfect (let alone ideal) and there will probably be a tiny bit of air somewhere there still, but after seeing the results it doesn't really matter:
On the first laptop (no separate GPU, just the APU), there was no problem getting the temps up to 80*C, and once it even had a thermal shutdown (90*C). Now I can't even get them above 65*C.
On the second laptop (APU as well as a dedicated GPU), the APU would generally stay below 75*C but the GPU could get to 85*C, and quickly. Now the GPU doesn't get above 70*C (with a 100MHz bump in core and memory clocks), and the APU doesn't get above 65*C (THAT'S WITH A 100% OVERCLOCK NOW TOO! 1.5GHz --> 3.0GHz).
It's been nearly 6 months since I did this and the temperatures are still that low.