Question RTX 3060 Hardware Issue or Possible Compatibility Issue on an older lower-end Computer

Jul 15, 2023
14
3
15
This is my first time making a post here (long-time lurker, first-time user), I have run into a somewhat convoluted issue with my new RTX 3060 GPU, that IMO has two possible explanations. A little over 5 years ago, I built my first PC, a lower-end entrance gaming PC. As such, I ended up making a couple of rookie mistakes such as cheating out on my PSU and getting a motherboard with little upgrade compatibility. Nevertheless, here are my specs:
  • i5-8400 CPU
  • MSI B360-A PRO
  • EV3A 500 Watt 80+ PSU
  • Gigabyte GTX 1060
  • 16 GB DDR4 2133 mHz RAM
  • (I also have an SSD, HDD, etc, but I didn't include these as they are probably not the issue)
About a year or two ago, I had my eyes on the shiny new RTX 3060, a substantial upgrade to my GTX 1060, which was starting to run into performance issues even on older titles. While I knew it was possible that I might run into compatibility issues with my motherboard, posts on here and other techie forums convinced me I was probably ok. As this was during the big GPU shortage, it was incredibly hard to find anyone selling even the RTX 3060 for around MSRP. Eventually, I found a very well-priced card on Newegg, however, it was sold by a VERY fishy company, Yeston. I did do some research on them, and they seemed to be a somewhat known brand in the East, but of course, I was, and still am, a little skeptical. I did decide to purchase back then, and I received the card in a timely matter, but due to other things going on in my life, I was unable to do the installation until about a week ago (Which was especially spurred on by the fact my other GPU was barely even able to run Hollow Knight, and required me to turn on V-sync in the Nvidia Control Panel just to make the experience playable).

When I finally installed the RTX 3060, things went quite smoothly for about a week (with the exception of this one session where my GPU reached temps around 80-90 *C). However, today, everything came crashing down when my PC just randomly blue-screened. At first, I wasn't sure what the issue was (and was praying it wasn't my GPU), but after booting it back up and immediately blue screening again, with a GPU error, it was definitely a GPU-related issue. As always, I checked for updates, fiddled about in device manager, updated drivers, uninstalled and reinstalled the GPU Graphics Drivers, and eventually used DDU in safe mode to attempt a clean install of my GPU. None of this worked and I even got greenlines when booting up into windows (without safe mode). I should also note, that for the few seconds before my computer blue screened, I managed to check temps, and they were all reasonable.

This leads me to my final conclusion. Either something is wrong with the actual GPU (Perhaps Yeston is a scammy third-rate refurbished GPU manufacturer who just made a bad GPU) or there is some compatibility issue. Despite what many on tech forms say, some blog posts or other websites such as this suggest that the RTX 3060 won't work with my motherboard's Chipset or CPU socket (LGA 1151), so perhaps that is the issue. Alternatively, perhaps the GPU requires more power than the PSU can supply, or a multitude of other reasons. Either way, I would really appreciate if someone is willing or able to provide context on this issue. If necessary I can provide more details or log files. Additionally, Yeston's warranty card is written completely in Chinese (which I don't speak) so I would love to avoid going through that hassle. Thanks for taking the time to read my long post, please let me know if you have any ideas.
 

Lutfij

Titan
Moderator
Welcome to the forums, newcomer!

EV3A 500 Watt 80+ PSU
How old is the PSU in your build? Try and replace the unit with something that's got less mileage under it's belt and is fairly new, sourced from a friend or neighbor to power your entire platform.
 
Jul 15, 2023
14
3
15
Thanks for responding so quickly, my PSU is about 5 years old, I got it when I initially built my PC. I also thought it might have been a PSU issue, but the crashes happened when the graphics card was not drawing that much power and the PC runs fine in safe mode or with the GPU turned off by Windows automatically (as in the graphics card in Device manager reads "Windows has stopped this device because it has reported problems. (code 43)." I might be able to get a friend to give my their PSU, but at that point it would be easier to just take the graphics card to their house and test it in their PC...
 
Jul 15, 2023
14
3
15
Should I disconnect my graphics card from my PC or keep my PC turned off to prevent damage while I'm working on fixing this issue? Also in case you want a copy of my "minidump" files, you can find them here.
 

boju

Titan
Ambassador
Seems like it's either the card or power supply. If motherboard wasn't compatible then the card wouldn't have worked when you first put it in. Motherboard isn't an issue anyway, oldish yes, but still modern enough. Heck, even older gen uefi motherboards can run latest cards.

Change psu first for something better, this needs doing anyway. I suggest Corsair Rm650x. Then go from there.
 
  • Like
Reactions: Contone
Jul 15, 2023
14
3
15
Seems like it's either the card or power supply. If motherboard wasn't compatible then the card wouldn't have worked when you first put it in. Motherboard isn't an issue anyway, oldish yes, but still modern enough. Heck, even older gen uefi motherboards can run latest cards.

Change psu first for something better, this needs doing anyway. I suggest Corsair Rm650x. Then go from there.
Thank you very much for the advice! Glad to know that the motherboard is a non-issue. In regards to your suggestion, should I attempt to test the GPU on a friend's machine? or should I just get the psu immediately and see if that fixes it?
 
Jul 15, 2023
14
3
15
The new PSU just arrived! (took a LOT longer than it said it would...) Hopefully, this will fix the problem and I won't have to deal with Yeston's warranty service. It'll probably take a little while to replace the PSU since my old one isn't modular so I'll have to unplug everything and completely redo my cable management. Crossing my fingers that this was the sole catalyst of the problem.
 

boju

Titan
Ambassador
It'll probably take a little while to replace the PSU since my old one isn't modular so I'll have to unplug everything and completely redo my cable management.

Blessing in disguise because there isn't pinout guidelines psu companies follow, so often pin diagrams change between them. That means modular psu cables aren't interchangeable between brands and there's high risk of damaging components if not 100% sure of cable compatibility. Easiest assurance is to replace every cable providing power.

Good luck with the new psu.
 
Jul 15, 2023
14
3
15
I just finished installing the new PSU, booted up my PC, and IT WORKS! I checked device manager and task manager, and both registered the GPU as working properly. However, I don't want to jinx it so do you know any good GPU benchmark/reliability test that I can run in order to ensure the GPU won't fail? I am already familiar with Cinebench as a reliable benchmark to test CPUs but what are the good ones for a GPU? Thank again Boju for all your help :)
 
Jul 15, 2023
14
3
15
Alright so an update, I ran 3DMark, the Assassins Creed Odyssey Benchmark, and played Hollow Knight with an uncapped frame rate. While both the 3DMark and AC tests went perfectly (the GPU performed as expected), when I booted up HK things started getting bad. At first things were fine, the GPU did go to 90-100% utilization with temps as high as 75 degrees (still acceptable though), but after a little while, I started encountering an audio stutter. At first this was all it was, but then the game started stuttering as well. Looking at task manager, CPU was spiking during these stutters so I just assumed the CPU was unable to keep up, but then my worst fears came true:
IMG_6183.jpg

So I guess I'm back to square one... Unless you guys can think of something else, it is probably a faulty card :(
 
Last edited:
Jul 15, 2023
14
3
15
Well that's actually possible, the fans were getting quite load so maybe they are overheating... I'll attach a diagram of my 'airflow'

image.png

I'll experiment with a fan, but how will I know if the issue is solved (since if the problem is inadequate cooling, I'll probably want to update the number, quality and placement of fans in my pc) as I won't want to leave the side panel open long term or it'll get really dusty...
 

boju

Titan
Ambassador
Hwinfo64 would display vrm info from motherboard if there's a thermistor sensor present, might not be.


If not then trial and error. If placing external fan helps then look at adding additional fans, another intake, to cover more of the motherboard.

What kind of cpu cooler do you have? A tower design with fans dragging air through would help.
 

boju

Titan
Ambassador
I mean motherboard vrm, voltage regulator module.

Your cooling looks sufficient, both cpu and other characteristics that would bring air flow over the motherboard. Was just poking a stick towards this possibly being vrm related, it was already a long shot.

You could try underclock the graphics card a little using Msi Afterburner, like dropping memory clock.