Showing results for 
Search instead for 
Did you mean: 

Drivers & Software

Adept I

Drivers Crashing/Black Screen on Rx 570

For almost a month now I was getting a Black screen seemingly randomly while using the computer after much troubleshooting I found that It happens when I launch certain applications or go on some websites on chrome. I disabled hardware acceleration on chrome which seems to have fixed crashing on chrome but still, it crashes when I launch certain games or play any video using vlc.

I did try reinstalling windows and that worked for like 2 days then it started black screening again, so I reinstalled windows again but this time it didn't do anything.

Also In the event viewer > system, it gives a distributedCOM(10016) warning. Tried some solutions from the web, didn't resolve it

Things I have tried:

  • reinstalling windows
  • system restore
  • disabling MPO
  • uninstalling drivers(using ddu) & installing older drivers
  • updating drivers
  • updating windows
  • Setting startup type to Automatic (Delayed Start) in Function Discovery Resource Provider
  • Disabling Distributed COM


  • Core i7 6700
  • Asus Rx570 4gb
  • 16 gb RAM 2133mhz
  • psu 550 watt
7 Replies

You might try what Deemon69 recommended in some thread and disable hardware acceleration through registry. You'll probably find instructions with google. I'm not at computer right now myself.


Yeah I tried that, sadly it didn't work, still black screens


550W PSU? When brand and model are not mentioned, I fear the worst. How old is you PSU and which brand and model is it?

Same thing with motherboard.

There is a reason for the saying "friends don't let friends buy bad PSU's", but you can extend that to motherboards. Almost always when someone has had issues with motherboards, they have either been made by some computer brand like HP, Dell, Acer, or alike, or they have been some budget models from Asrock, Abit or Gigabyte (or some lesser known brand).

You could try unplugging everything, leaving only one DIMM, disabling XMP/DOCP and underclocking CPU and GPU for testing purposes. If problem is solved, then bring them back to normal speed one at a time to find the culprit.

And of course temps, what Deemon mentioned.

Thanks For The Suggestion!!

I tried installing the oldest driver I could find on the AMD website and that seems to have fixed the issue, however, I fear it may just be temporary.

I think the PSU is the main culprit, its a FSP Hexa+, I know that sounds like a odd no name brand & I brought it second hand (I had very little knowledge of parts)

and as for motherboard, I have Asus h110m

If the problem comes again, I will try your method and maybe look for a new PSU


Good day!

Is there an overheating of the equipment? You probably already noticed that the same thing happens in all cases with the PC shortly before the black screen error occurs and at the moment the black screen appears, could you tell? At these moments, there are no strange flickering or short-term disappearance of the image on the monitor, short-term freezes of the picture, etc.?

Nah, nothing like that I had overheating issues before due to fans not working but this was different it just went black screen no flicker and temps were normal.

But the issue is resolved for now by installing older drivers.

Adept I

Hello, my name is Dr. Meklin and I have used the  RX570 since 2019 and had the same issues peoples years later are posting. I did not fix them until 2024. It is not a ryzen problem, amd problem, or bios problem, or am4/b550+ chipset problem. How to fix the RX 570:
Take it apart, get your thermal paste, take off all the thermal pads and wear a latex glove after you get them off. Roll them all into a ball, it was years before I did this and mine was like hardened baked clay. You will now have a hard ball with chunks falling, put some thermal paste in it and rub it on every surface of every chunk and squeeze them together, fold them like steel until you have a perfect blob or soft play doh which is going to get hard later. roll out little tube shaped turds or whatever and put them all back on but down press down, the backplate will do that. i never really had a problem with the main thermal paste but redo it often probably 10 times since 2019. that isnt the part that is getting to hot, the VRAM is. i dont believe there is a way to detect it. drill a hole where the fans inside connect to each other, because you want that wire going out. if you want to do this your own way instead of my way, run one of the fan lines directly to the motherboard. this is how i fixed the rx570, by reducing the power draw. i have maxed out OCCT on overclock mode and drawn 250+ watts from this card, more than it probably can in standard setup. for my setup, one power line goes to the gpu, as it will not post without this (rx570 sapphire 8gb), and the other line goes to a MOLEX to 12v and 0.7v adapter. i just found it in my house but you can order one somewhere, it has two lines coming off the adapter once it makes connection to the molex, and i use the 12v line to hit my gpu fan. it maxes out that fan as it thinks it is a chassis fan and is detected. i played on this for a while and no crashes but eventually i got one even on standard tuning profile in radeon settings. at this point i realized i had achieved some stability, but i didnt put the backplate back on, nor modify it yet at this point with new paste and a fan. i have only drilled a hole in the bottom of the card and run the fan power linesout the bottom of the card where they split, and this has only reduced the power draw from the gpu's own fan power jumpers, nothing more.

so i realize i can overclock the card and play for a while, but i see an amd64 vega gpu crashing and he put intel fans on his gpu near the vram thermals. i could have fixed maybe just by putting the back plate on, but at this point, i knew i was truly facing a stability issue, not bios, not anything, i look at OCCT and see how fast a cpu heats up, my ryzen 3600x from 40c to 80c in 2 seconds on the 100% cpu max out test, and assume that is happening to the gpu and it really, really needs to be cool. see more posts about stability achieved with undervolt and max gpu fan 100%, so i set that as a staple. my 'chassis' intel i5 fan without a heat sink just hanging off the side of the vram area, almost touching my gpu fan cuz theres no room, taped on, i reccomend just drill a few small holes the same size as a motherboards how a cpu cooler mounts into that u can easily just let the fan 'sit' on top in those holes so it doesnt move it doesnt need to lock in at all.

i will make a video of my blasting out this card, and probably even show what happens if i unplug the third gpu fan which plugs DIRECTLY into my cpu chassis area and is maxed out 100% in my bios settings. this is a very very good gpu and the technology at that time wasnt fully understood by the 3rd party companies tweaking them in a 3 celcius chinese basement factory they assume we will get better cooling maybe like liquid. anyways enjoying my 40+ fps and oh yes i do notice when in dota 12 vs 12 and there are 24 ultimate spells going off at once, the very moment the gpu would have crashed with a small stutter that any modern computing system would experience when 48 sunstrikes go off at once, but it remains stable. others wont admit their placebo fix for this card were failures, they dont come back to the forum years later to help their poor brethren still using this card, their moms got them 2080ti's by now. but dr meklin is here and i have fixed this 4+ year issue, finally. be sure to repost this and check out the video on my which will be up soon but you have enough here to visualize what you need to do to get this AWESOME (and i am literally a top 50 rust player) card working stable for probably a decade to come