[Need Help] Extremely Hot GS65 i7-9750H (GenTechPC)

I don't believe 92-93C under load is normal with liquid metal, you can find many examples of people having lower temps after applying liquid metal and not only on this model. Personally I would send it back no matter what the reseller says, but it is up to you. Before you send it back, can you run AIDA64 stress test while monitoring CPU temps with hwinfo? I am curious to see how hot CPU gets, especially each core.

It may be "common" due to very bad thermal paste application or defective heatsink but it does not mean a person should tolerate it ;-) Here are the temperatures from my old MSI laptop. It is not the same model but it does have 8750h CPU, so same amount of cores. When it arrived, there was huge core temp variation and just hot temperatures in general due to very bad stock paste job. After I personally applied better thermal paste (IC Diamond - it does leave scratches but I did not have better thermal paste at that time, only stuff like MX-4 which is NOT acceptable for laptops because it is NOT THICK ENOUGH) here are the temps when running "Auto" fans and without undervolt, running AIDA64 CPU stress test. Note the temp difference between cores.

 

Thank you for such a detailed reply. Sure, happy to run AIDA64 - can you let me know which version would be best? I saw there's the Extreme, Engineer, and Business. If you let me know which one and which settings to apply, I'll run it for you. Also do you want me to keep the laptop undervolted (currently at -140mV) while running the tests?

I think I can get one of my freelancers to take over the video editing for the next couple weeks because I think I should also send it back and get liquid metal applied. I'm just not happy with the temps and while Ken@GenTechPC said some things I was suspicious of, he has offered me great customer service so far. I just wish they would have contacted me about the thermal paste if it wasn't a good choice for the laptop, especially since I waited the full 10 business days for it to be shipped (and had called them asking about the wait time).

 

The Extreme AIDA64 would be enough. It will run in free trial mode forever. After running it, you can go to Tools menu then Sytem Stability Test. You can uncheck everything except "Stress CPU". Then just run it for a few minutes and also run hwinfo with it. You should not be seeing more than 10C difference between CPU cores if thermal paste was properly applied and heatsink is not defective.

Here is a Reddit post which shows result of applying liquid metal using same model as you have but different generation of CPU:
https://www.reddit.com/r/MSILaptops..._stealth_thin_result_of_repaste_and_applying/ this includes undervolt but this is still a very impressive difference, especially since that person says that he or she used to see up to 90C during same benchmark.

 

Perfect, thank you for telling me all of this. Am I reading that right - they are seeing 68C under load? Here are the results of the test after 5 minutes: https://i.imgur.com/qU6dYm5.png

Edit: Here's when it's running also, if it makes a difference: https://i.imgur.com/76P1jwE.png

It seems Core 1, 3 and 5 are really a lot cooler than the others. Not a full 10C cooler on all of them, but some of them.

 

Yea, it is clear that Kryonaut is definitely wrong for this laptop. I would send it back for repaste with either a thicker paste like IC Diamond or Gelid GC Extreme or a liquid metal. And you can at least see what kind of results to expect from liquid metal application.

 

Much appreciated. I'll reach out to Ken and see what the logistics are on sending it back, he said that the offer to re-paste stands indefinitely which was very nice of him to offer. Do you think that the core temp differences are from a pasting or heatsink seating problem?

 

It could be both the bad paste job and uneven heatsink surface or uneven heatsink pressure. In any case, using thicker thermal paste or liquid metal should solve that stuff. I do not have liquid metal but this is what I had on my old laptop before I repasted it:

 

I wouldn't use liquid metal if it indeed is a heatsink gap issue.

 

How would I be able to tell if it's a heatsink gap issue, though @custom90gt ? Should I just send it back to Ken@GenTechPC and have him take a look at it?

 

Really the only way to do it would be to remove the heatsink and check the spread. I know you're not keen on taking the MB out yourself, and I don't blame you.