Hello, I have a problem with my Sapphire RX 590 Nitro+ graphics card... The graphics card had high temperatures, so I disassembled it... I replaced the Thermal pad and Thermal paste... When working with Excel, the temperature is at most 40 degrees.... but when I turn on the PUBG game, the temperature rises to 81 degrees and the fans speed to 3200RPM.... Sometimes it happens that the game freezes and I have to hard shutdown the computer. My graphics card driver is updated. The graphics card is not overclocked. Thanks for help
ARCTIC MX-4 Thermal Compound
ARCTIC Thermal Pad
PUBG on HIGH Settings
Solved! Go to Solution.
check if your fan curve is always near 100% in adrenalin
you have to check manual settings to see
maybe some pads are to thick or too thin ? or problem while remounting ?
are your temps worse than before disasembling and replacing ?
Before the change, the temperatures were a little worse. Currently after shutting down the game it shows -
fan speed on the graphics card - 3051RPM
Temperature - 35 degrees
Consumption - 38W
Graphics card used at 22%.
I don't understand what is wrong.
check if your fan curve is always near 100% in adrenalin
you have to check manual settings to see
That poor card deserved a third fan to keep the temps under control.....
But I'm more inclined to a bad assembly, I'm sorry.
I learned this from hardware unboxed, try laying down the PC so the card stays vertically and (very) gently push the heat sink towards the backplate, if temps get considerably lower, then its bad assembly.
Like always, do it at your own risk!
Pressed and un-pressed from a bad assembly, the RoG Strix 5700XT had this issue from factory...
When I set the fan settings manually in Adrenaline SW, everything is fine... After turning off the PUBG game, the fans automatically reduce the temperature. If I did the assembly wrong, the temperatures on the graphics card would be high from the start... am I right? well, I have a stable 35-40 degrees on the graphics when starting the PC and surfing the Internet....
I updated the bios. Maybe that helped too. Anyway, everything is fine now.
Thanks for help
Good to know that things got back to normal.
Another good piece of advice is:
Undervolt the GPU
i don't think i is good to adivse to tweak and add possible instability causes..; there seem to be no temps issue here apparently .., and OP doesn't seem to have great experience in tweaking .. your advise is good for people knowing what they do.. (i know people having problems accusing drivers as they are oc or uv and don't understand what it implies), so until you want more people coming on this forum lol..)
I'm very sorry but that is not up to you to decide.
The advice is valid and its well bellow the skill level required of a GPU repaste. Which He did.
The user can take my info and decide for Himself if its valid for His use case scenario, undervolting can help substancially with temperatures.
The newer cards even have auto undervolt option available at adrenalin level.
I had my RX480 undervolted for quite some years and less voltage plus less temperature will prolong silicon life.
i just warn OP, not trying to start a fight with you as you seem to think , ..and i thought i wouldn't lol .. if you have little experience with auto uv or oc you know this doesn't implies 100% stability in all cases or it would have been a standard always activated parameter
Don't worry, I'm not.
The apologize in the beginning of my post was sincere.
Although.. we have to avail what the user did, so if He already embraced the "adventure" of a repasting, a "project" far more risky and if done badly it means a dead card on His hands, plus His system already below the 100% stable mark. Then there is very little (or nothing) to lose.
Don't forget that this community embraces overclocking and modding.
It has its risks but so does repasting.
Don't forget to deinstall MSI Center if you are using a MSI board.
It fiddles with the GPU setting every restart if it is installed and the plugins for GPU are installed. !Even if you don't have it in "Autostart" .
It resulted on my 590 Fatboy in Screen being full of pink blocks like damaged vram (every ~10th pc restart) which couldnt be fixed by a windows reboot. I had to unplug the PSU. Some would think the Graphicscard is dead. But it runs normal since i deinstalled MSI Center over a year ago.
And check if you had the same issues with the 2nd bios, had 1 send to me for repair were the bios (customer said: never touched it) got corrupted and fans wern't running. After a reflash it worked normal again.
And yes the 590s are definetly bad designed for heat transfer, especially to block vram overclocking.