Weird temperature issue after a repaste

I have an alienware 17r4 with i7 6280hk and GTX 1080. As people here know, the cooling system on this thing isn't exactly the best. It was pretty fine at the start of usage where temperatures on the GPU would mostly hang around 60 degrees, go up to 80 when stressed. The issue started getting worse with time and so I decided to fix the issue by repasting and repading the laptop (btw. dell toothpaste stamp was mind-boggling when i saw it). The pads also exchanged and the paste was reapplied using Thermal Grizzly Kryonaut and Fujipoly 17 W pads. Everything was done according to a detailed guide (iunlock) plus pads were checked for coverage and actually making proper contact with each part of the heating system (one pad that was 1mm was not making contact at all so I used a 1.5 pad).

The biggest problem was always with the cpu and that was fixed, I can easily cruise at higher clock speeds without enormous temperature spikes and throttling every other minute. When I start playing games however, my GPU temperature goes up to 90 degrees! The laptop feels cooler to the touch, I don't get thermal throttling that was the case before after reaching 84 degrees on both cpu and gpu, but the temperature of the gpu is 89-90 degrees constantly! Even when lowering the resolution from 1440p to 1080p and lowering the quality of the graphics the temperature does not drop. Game settings are set to adaptive so it is not the maximum performance issue. The weirdest thing is that the laptop feels smoother and better in games overall, like it has enhanced performance.

Did anyone ever encounter anything like this? What do you recommend me to do? Is 90 degrees celsius safe for a notebook GTX 1080? Thank You for any input

 

Maybe cleanup your fans, and try to rebalance your heatsink. 90C spikes are not bad, but contstant load will eventually wear down your GPU, and GPU fan. After that parts like the PCH and SSD drives will likely faill due temperature..

Have you applied enough thermal paste ? Did you check the coverage/spread of the paste ? Is it instantly or does it rise slowly to 90C.

If you can answer those that could paint a more clear pictures. Log with HWinfo and post the log with all the gpu info. Maybe that'll paint clear picture.

 

I took the paste, got a pea sized dot on the middle and very little dots on the corners, then spread it with the spatula. With first seating of the heat sink I had to repad one thing so I took it off again and the paste was spread on the heat sink and the die of the GPU. The die on the GPU had a little bit of excess paste that gathered in lines of a little thicker thermal compound on the die so I think it should be enough because there was a little excess but it was a thin layer overall.

When repasting yesterday, everything was properly cleaned. Idle temp hangs around 40 degrees (room temperature is around 25 at this time of the year). When I get to the menu of Dark Souls 3 for example it gets to about 44 degrees but as soon as I get into the game itself, the temperature steadily and rather quickly raises to 90 degrees (in about 30 seconds to a minute) and just stays there with minimal fluctuations (86-91) but stays mostly at 90 degrees.

 

Alright can you show us the temperatures on idle with HWinfo (gpu and CPU), then another where you ONLY stresstest your CPU with XTU or something (again both GPU and CPU). I have an idea whats going on,

 

Good, so the paste is spread good. What about the gpu fan speed, and cpu temperature. Have you at what speed is your GPU running and how much watt does it pull ?

 

I have my CPU at 38x and in OCCT it stays around 72-78 degrees. When gaming it is around 80 (probably getting the temperature from the GPU) sometimes spikes to 86 but that is for a split second, normally it hangs around 74-80 degrees. I set up the fan to be blowing 100% at all times to check, but it changes nothing so even without it, it properly kicks in. I'm gonna check everything with a friend that I was repasting it with (he has much more hardware experience). At maximum temp on the gpu it is drawing around 171W.

 

Could you provide those 2 screens I was requesting? would give me much more insight on the problem.

 

https://imgur.com/a/k5Uytgj

Both only after a couple of minutes. Though the problem is, that the stress test in OCCT lowered my clock speeds to 3.2 Mhz after an initial spike to 3.8 Mhz. Where in games it is mostly 3.8 Mhz so the temperatures go up compared to this. If you need me to run anything else, hit me up. Also if you have an idea, you can tell me now what to check and look for coz I will be taking the pc apart in the next hour.

 

Hmm you would expect from what you describe that the the pads that you have replaced around the GPU might be a tad too thick.

 

I have the supsission that the heatpipes actually warm your GPU up. It transfers the heat from the CPU to the GPU, which is kinda weird since normally on stock it's the other way around.

 

Any fix you can think of? It wasn't a problem before the repaste. Any specific pads on any screenshot you can point out could coz this issue? should I exchange them with thermal paste or thinner pads?

 

You can try and repaste again and see if it does anytyhing.

 

If you're using kryonaut then I suggest you should try another paste. I used kryonaut before and get much higher temp even when idle (65-70). Then I tried mx4 and my idle temps stay at 50-55°C.

 

I just ordered electrical tape and conductonaut. I am just gonna go with LM and I think it will give me the results I am looking for and being extremely careful with it should be just fine not to short out the motherboard. I'm also gonna check which pads are too thick and replace them with 0.5 ones. Cheers guys, thanks for all your input

 

Fujipoly 17 pads do not compress enough. You need gelid 12W/mK thermal pads that will compress so you make proper contact. Especially for LM. Don't change the cpu ones unless you really compress a .5mm pad.