fallout_2077

Oops, forgot to mention those. The temps for both CPU and GPU sit between 70 to 72C while in game. While compiling source code in Linux the CPU temp climbs to 90C and stays at that temp.

I posted here a couple weeks ago regarding my MSI GT70 2PC Dominator shutting down when playing demanding games such as Elite Dangerous and Cities Skylines. I thought the issue lied in the broken CPU heatsink mount that's affixed to the motherboard or faulty sensors, but after buying a new motherboard, the problem persists. My only guesses are the problem is either with the CPU or the GPU. Is there any way I can narrow down the problem to one of these two components? Or do you folks have any suggestions as to where else the issue could be appearing from? I've ran diagnostic tests on the RAM, and all is well on that front. Thank you for your time.

I bought the standard JB Weld that is shown in the vid (I bought it and applied it before I saw your response). It didn't work... It hardened, but the epoxy didn't bond well to the surface of the PCB. I bought a soldering iron to attempt that route, but I made another thread for advice on that before I travel down that road.

I made a thread nearly a week ago regarding the heatsink mount that separated from the motherboard in my MSI GT70 2PC. I tried epoxy, which failed to resolved the issue. My other option now is to attempt to solder it back to the board, which judging by the other mounts for the CPU and GPU, is closer to how they were originally attached. The thing is, I'm clueless as to how to tackle this, as this would be my first time soldering. I bought a soldering iron to practice on other electronics with before I mess with my laptop. Any advice would be greatly appreciated!

I found this vid showing JB Weld for electronics repair. Would this work to solve my problem?

Okay then, thanks for the advice! Is there any special kind of epoxy I should be on the look out for, or will any epoxy work? I've never worked with the stuff, so I know absolutely nothing about it.

Here are more images of the laptop, detailing both the mount and the areas around the CPU & GPU to show what's going on those areas, including the condition and placement of the thermal pads. I uploaded these to Dropbox and shared the folder so they'd be easier to see. https://www.dropbox.com/sh/vudn8qm8s8zuvjt/AAAQaOqDE-9GhjLAEn_nJ4sra?dl=0

It's a metal mount, and below it is a "hole". I'll have to take a closer look at the area again and see if there's any possible way for any epoxy to leak through to the other side of the board. If it did happen to leak, can epoxy damage the board? I judged that those mounts were soldered onto the board. I'll take the laptop apart shortly and snap some better photos.

I had that thermal paste lying around for a few years... I bought it when I had to replace and upgrade the GPU in the laptop when the original failed. I knew that paste expired, I just didn't think it would expire so soon, though the runniness was a pretty dead giveaway that it was starting to show its age. So do you believe that the paste is a pretty strong possibility? As I stated above, the average temps during gaming and intensive tasks is 70C, with a max of 80C (the CPU gets a lot hotter than the GPU during these tasks). Could most, if not all, of my issues lie in the thermal compound, then? Should I even worry about fixing that mount?

My laptop fan sounds perfectly normal, with fan speeds well within the expected range while under load and during idle. And aside from that one mount that separated from the board/chassis, the other two mounts are fastened tightly to the heatsink. I had considered the thermal paste as a possible culprit, as sometimes when squeezing some out their is an initial bit of runniness, forcing me to either "pump" the plunger in an attempt to mix it within the tube or to mix in when it is out of the tube and on the chip. And their are one or two pads on both the GPU and CPU heatsinks that are split apart with a slight space. I may have some extra pads lying around, hopefully. And I bought a brand new power brick last week as my google searching led me to believe that power could be the issue, though following replacement the problem persists. At the moment my suspicions lie on that one separated mount. Here are some images... an image is worth a thousand words, after all.

For the past couple weeks I have been experiencing predictable shut downs while playing demanding games and performing intensive tasks, the two prime examples being Blender 2.8 and Elite Dangerous (within minutes of loading a save). Until last night, I had presumed that all of those crashes were isolated to Windows, as I was able to run Blender and play games just fine in Linux... until last night, when I saw my first shut-down while playing Elite Dangerous. Linux can apparently handle Blender fine, but Elite failed at the same point as it does within Windows. The temps never exceed 70C for GPU and CPU, which make me think that if it is a overheating issue, it's more of a sudden thermal spike causing an overheating shut down. I experienced the exact same shut downs a few years ago with GTA V, but the problem was solved when I lowered the min/max CPU states in the power plan settings. The same can not be said this time around, however... I lowered the min/max states all the way to 60/65 respectively, with the shut downs ever present. Before I was experiencing the predictable stream of shut downs, I was experiencing BSODs relating to memory management; these BSODs never happened during gaming and usually only happened shortly after powering the system on or while shutting it down, and usually only once a day. However, after running Windows' Memory Diagnostic Tool, no problems were found, and after several reseats of the RAM sticks, the problem went away. Within a day or so of clearing that error, those BSODs were replaced with predictable, albeit worse instantaneous shut downs, though about 10% of the time the system actually restarts like it's supposed to do. I ran MemTest86 overnight and it did not report any errors, so if there is something wrong with my RAM, Windows Memory Diagnostic Tool and MemTest86 are unable to find it. I'm dumbfounded as to what the culprit could be; faulty sensors, a failing motherboard, CPU, or GPU, or is it my disconnected CPU heatsink-to-motherboard mount? I'm at a complete loss, so whatever you fine, intelligent folks need to solve, or at least point me in the right direction of, my problem, please let me know and I'll gladly provide you with any and all info you need regarding the laptop.