
UNSOLVED
Closed
Adventures of an m15 R1 owner
So, another day with the M15, and another story to tell.
Up until yesterday, I hadn't really been paying that much attention to the cpu temperatures of my M15. Out of curiosity, last night I left HWInfo running while I had a blast through Doom Arcade mode for a bit.
When I'd done, I checked on the reported temperatures, and sort of as expected, saw that I'd hit really high 90's (97/98), and that it had been thermal throttling.
Interestingly though, only 3 of the 6 cores reported temperatures this high (cores 0, 2 and 4 on HWInfo). The other 3 cores were a fair bit less (high 80's if I remember correctly).
Tried a few other things, and noticed that the same 3 cores hit higher max temperatures that the other three. There was between 9-10 degrees difference consistently.
So this morning, I decided to take the plunge and do a repaste as I thought the temperature difference might be a sign of issues with the factory job. Tested a few more games, and while the overall temps may have been a few degrees lower, the 3 cores were still hitting mid 90's, with the other 3 cores again in showing in the region of 10 degrees lower max temps.
I then decided to give the stress test in Intel XTU a try, and this revealed a similar situation with half the cores about 10 degrees lower max temp. The three highest 'problem' cores all hit high 90's again and showed thermal throttling.
So after further reading, I decided to try an undervolt. I started off with a fairly conservative -0.100 undervolt, and this gave immediate benefits with the stress test in XTU. I was no longer getting thermal throttling, and all cores were running at 3.9GHz.
Ended up increasing the undervolt a little more, finally settling at -0.120.
Ran a few more stress tests, and games, and the highest temperature an individual core hit was 83 degrees, which is pretty decent in my book.
There's still a considerable difference between the max temps of the 6 cores, so this maybe highlights a physical issue with the seating of the heatsink. Although it is only the max core temp that varies like this - the average temp of the cores over the period I was running HWInfo were all within 3 degrees of each other.
Not sure how this will format, but this was the view in HWInfo after running stress tests and games with the repaste + undervolt (the figures are temperatures for Current Time - Maximum - Average:
- Core 0 -- 43 - 83 - 54
- Core 1 -- 44 - 73 - 51
- Core 2 -- 43 - 82 - 53
- Core 3 -- 43 - 73 - 51
- Core 4 -- 44 - 79 - 53
- Core 5 -- 42 - 71 - 50
I don't know how the cores are physically laid out on the processor, though I guess it isn't beyond reason that cores 0, 2 and 4 are on one edge, and 1, 3 and 5 on the other, and would possibly show that the heat sink doesn't have the same pressure all over the cpu. Same temperature differences with the factory job and my repaste hopefully show that the paste itself isn't the issue.
Overall, the temperature difference is curious, but I'm really pleased at the overall resulting temperatures I'm now getting, and just thought this may be useful info for others.
Back to some demon slaying!
This conversation is no longer open for comments or replies and is no longer visible to community members.
Responses (1)
Solutions (0)

Whokam
20 Posts
715
1
Posted December 26th, 2018 07:00
Following the repaste and under-volting that I'd recently done, my M15 had been running like a champ for several days. Stress tests and gaming were only peaking core temperatures of around 86 degrees, and i wasn't seeing any thermal throttling.
Then last night, under the blue glow of my keyboard as I had a go on Forza Horizon 3 that had been downloading for the past 2 days (thank you MS 'Play Anywhere', but also thank you middle of UK internet speeds...), I heard a blustering noise. The fans had quickly ramped up to full, and temperatures in the room started rising.
My heart sank as I alt tabbed out, loaded up HWInfo and saw core temperatures of 100 degrees and thermal throttling. Closed Forza, tried Doom and saw the same. The stress test in XTU immediately saw temperatures of 100 degrees as it was started (and quickly stopped!).
Even though the laptop had been switched on for a couple of days, the under-volt seemed to have come off.
So, with XTU fired up, the under-volt was reapplied. But what was this... still high temperatures and thermal throttling. This wasn't making any sense. HWInfo was showing the under-volt, but temperatures were still in a range that seemed to indicate that it wasn't taking properly. After a few restarts, and total shutdowns and power ons, the situation was the same.
As this was all after a long day of festivities, I decided to shut the laptop down and go to sleep.
So, back to reality this morning. I fired up XTU and reapplied the under-volt, but the stress test still showed temps hitting 100 and thermal throttling, mainly just on core 2 which seemed strange. Everything had been fine for a few days, and now the 3 hot cores had gotten hotter, with them hitting 100 degrees fairly regularly, and the other cores also showing a temperature increase. I had also observed that right after start up, as Windows is still doing it's thing, the cpu would report 100 degrees with nothing else running.
Have I fallen foul to the various issues that seem to have been plaguing the M15... unclear.
Not wanting to give up, I opened the laptop up to give the thermal paste a once over. Oddly, it did seem as though the screws round the CPU had become a little bit looser. There was room to tighten them up every so slightly.
Had the thermal pads in this area become a bit more compressed and loosened the pressure up or something? Unclear.
Removed the heat sink and inspected the thermal paste, and to my eyes it all looked ok. Both CPU and GPU were showing evidence of good coverage. Cleaned it up, and reapplied fresh paste to ensure good coverage once again. Started windows up and XTU, ensure the under-volt was applied and ran the stress test. The same hot cores still showed the 10 degrees difference in max temperature between them and the other 3 cores (behaviour that I had seen in my M15 since the start), but now, max temperature of the hot cores was 80 degrees! Few repeat stress tests, and a run through Doom arcade mode, and the max temp I saw was 82 degrees.
So ultimately, I'm still not sure what to make of this. There definitely seems to be some kind of issue with the heat sink mount that is resulting in uneven thermal performance on the CPU. The thermal pad on the heat sink that covers what I think are the large grey VRM's looks fairly thick, and the VRM's leave quite a deep imprint in it where they have compressed it. Could this be a potential source of uneven pressure?
Either way, I'm now back to what I would consider 'good' temperatures for a thin and light machine. Still unclear what caused temperatures to increase again, so I'll have to keep a close eye on it.
The Thermals have been defeated once again.... but will they return..