Showing results for 
Search instead for 
Did you mean: 

Graphics Cards

Journeyman III

590 Fat boy crashing to black screen

When playing some games at random times my GPU will max out its fans then hard crash to a black screen, PC still runs tho. When checking temperatures im at around 75c while gaming. This issue does not happen when im not gaming. I have tried to turn off 0-rpm fan and increase the power consumption it can take. I have had the card for about 2 years now and never really seen massive issues until (at least I think) I installed some AMD drivers. Does anyone have some feedback/input that I can try to fix my issue? Any help would be appreciated, thanks!

Processor: AMD Ryzen 7 3700X 8-Core Processor 3.60 GHz
RAM: 32.0 GB
OS: Windows 11 pro X64
Driver Version: 30.0.15002.1004
AMD Software Version: 22.3.1

GPU: XFX 590 Fat Boy

5 Replies

Even though GPU temp is not going further than 75°C, I would monitor the VRAM if possible. You can download HWInfo and use their sensors to check VRAM temp. If it is not going hotter than 75°C either, I would try a slight decrease in clocks (GPU and VRAM) using MSI Afterburner or something similar.

RAMs could also be faulty. Try MemTest64.

I would also try Windows 10, in order to avoid any type of Windows 11 compatibility problem.

If a clean install of Win 10 doesnt work, I would try a different PSU.

Sorry for the late reply (never got a notification you responded). All my memory tests passed, and temps maxed out at 168.8F. After doing some investigation I have noticed that cooling could possibly be a major factor here. I pulled my card out of my system, I started checking the screws for the GPU cooler and they was just finger tight not really applying pressure to the VRAM thermal pads or the actual chip itself. I pulled the cooler off and applied some new thermal paste and thermal pads for the VRAM and changed my fan curves. I did see some strange areas near the VRAM which could mean they may have overheated at some point (check image attached). I put the cooler back on and tightened it tighter than what it was. I did not have any crashes but temps was still fairly high. What my current assumption is since I never actually checked VRAM temps before I opend the ticket im going with they was possibly overheating, if i have any more crashes I will lower clocks for both. I have seen other articles saying that the card had a weak cooler as it is. Do you think this would be a valid assumption? Screenshot 2022-04-03 201816.png20220328_161907.jpg


Those black paste things are solid, right?

They are part of the design of several boards. I usually see them in red. However, it´s not a sign of overheating.

If temps don´t drop after that, you should check that the vram thermal pads are making contact, and, if necessary, change them.

There are several heights of thermal pads, for example you can be 1mm, 2mm, and a lot in between.

Also, forcing the coolers to run faster should help you.


When the Original fans on old r9 390 died. I had initially had a hard time finding exact replacements ( i seem those xfx fans all the time now of course) but I had 2 120 mil Noctua's that I zip tied to the heat sink, i did change the thermal paste to arctic silver as well. and I honestly think it gave that card several more years, and phenomenal temps compared to before, best part It was still roughly the same dimensions as it was with stock fans and orignal shroud. Maybe as well as new paste and pads, higher quality fans. I understand those 590s are hard to cool. The 390s where pretty hot too. if i reacall she was usually 71 c at full load and after that rig job it was like 62c and it was the end of the summer. Its hot where I live in Idaho at that time of year. 

DId you check for bios update on your board as well? I had that issue too on a acer board. Bad PSU will intermitantly do this sometimes for months until it finally goes belly up. Is the crash consistent, I mean does it always have the same set of circumstances:? for instance Under heavy use rendering? or does it happen watching youtube as well? good luck to you

Ryzen 5 3600 &--Crucial Ballistix 16GB--GTX 1080ti

About 3 weeks ago I updated my bios. The crashing only seems to happen at random times while rendering, few examples would be, Ready Or Not, Farcry 5. I do not seem to have any crashing issues while preforming non demanding tasks such as, YouTube, coding and so on. Crashes are fairly inconsistent, power supply is a ASUS Rog Thor 1200 watt which I have hooked up a tester and it also passed. As of yesterday I did not seem to have any crashing issues while running my GPU threw stress tests or playing the games that roughly 85% of the time result in a crash.

I currently have the stock XFX cooler, so I may attempt to do something similar to what you stated.