Guys I have a temporary and permanent solution.
given you dont have any problem with psu, and memory and some other hardware and you are sure that the gpu or driver is the problem try the method below and get back to me.
And please tell me if it is work for you or not so we can help each other.
Sorry for bad english or wrong grammar. English is not my native language.
Go to performance then tuning.
Adjust your freq state 6 and 7 to
BASE CLOCK "its depend on amd card"
Search your gpu specs and you will see it
The easy way is, take the slide clock freq to
" -8 to -11 " make your way from -11 much more stable to -8 "sweet spot"
Take not the lower number -11 is much stable than -8.
If you are stable on higher number posible -8, and make a rigorous test.
Its not gonna hurt to test it to higher number -7, -6, and so on untill you make iit crash again,
Take note the lower its NEGATIVE number the higher the value. If you confuse when i say -8 is higher than -11.
Ok , obviously 0 is the default value.
As you can see amd gpu have boosting features, its like an over clock that's why its unstable.
You can also include it with aggressive fan curve, but first test it with all default value other than freq. Slider.
I alway crash on SCAVENGERS,
Temp is not a problem.
60 to 72c, while playing.
So far I test -6 on freq, and sometimes it crash, but this time i can play 2-3 games onscavengers before its crash, in deafault setting i will insta crash 2-5 mins in game.
Now i test -11 i play a lots of warzone and scavengers around 10pm to 2am without a single crash.
If your card is not on warranty and willing to take a risk.
Another thing is replacing the thermal pad and thermal paste, as reported from others they replace thier thermal paste and pad and now working properly and with the default setting , boost or auto oc even though its not officially state that boost is actually oc.
Oc mode is much higher oc as far as i understand it.
I did not replace my thermal pad and paste as of now. Im satisfied and for now i dont have time and items to do the job.
Reason behind this thermal pad or paste is that by factory sometimes the VRAM of gpu is not properly cooling, we cant see it to gpu sensor , the temp we see is on gpu core not its vram.
Its all base on my research and as i said above i did not replace my thermal paste or pad as lowering my freq from -6 to -11 solve my crashing issue for 4 hours i dont know if its hot like in the morning, but so far, 4 hours of playing give me a good result. Cod and the notorious crashing game in my case, scavengers.
Before you replace your thermal pad. Or paste be sure you have a proper pad thickness different gpu have diffirent pad thickness make sure its a quality one.
Sorry for a very long post, but i hope i helped 1 or 2 people out there, it will.make me happy!
Please report back if you try these so we can help someone. ☺️
Now for the finally!!!
Sorry to say but we should buy NVIDIA and avoid AMD Cards,
Yes its cheap and powerful vs their competitor but its really unstable,
I hope i helped someone.
Please update me!!
I installed Auto-Detect and Install Updates for Radeon™ Graphics and Ryzen™ Chipset Drivers for Windows at https://www.amd.com/en/support
I ran the repair/reinstall option and so far I haven't had any timeouts for the last few days.
Same here with RX580...
I think it started on 21.3.1, can't even remember.
AMD known Problem list for AMD 21.3.1 "Radeon RX 400 and 500 series graphics products may experience a TDR during extended periods of video playback."; 21.3.2 = same.
Then AMD removed it from the known problem list on later updates, but never shown up in the Fixed list either, so it's like AMD just removed it from this planet and didn't also fix it...
After updating here like 3 or 4 updates ago it started the same problem. My PC just keep crashing on firefox web browser randomly, almost all games crash and I updated every time they made a update after and I'm currently on last update and nothing fixed it, just get less crashes on Firefox and are faster to recover and almost all games keep crashing the same way making it impossible to use this GPU broken by AMD again for like the hundred time since I bought it...
Already used DDU on safe mod and remove all AMD stuff, already tried the full clean install from AMD driver too, same thing, nothing fixes it...
Thats all a crap...
I never updated windows 10 for a **bleep** long time, so the problem has nothing to do with windows. The problem only started when AMD broke the drivers like 3 or 4 updates behind and they even had it in the known problem list, but somehow on recent updates it's not listed anymore and never went to the fixed list...
Then who uses AMD utilities? I guess we all use the DDU in safe mode for many years and before it we had other...
My Drivers are always on default settings, never even used gaming mode, I have AMD Freesync and all features OFF.
And the only solution is to roll back a few updates ago and loose all the latest gaming performance improvements sadly.
problem is the windows and AMD drivers we did not had this issue until windows 10 start to update and add more **bleep** that we do not need for gaming. lot of them taking our Ram and HDD and CPU . also windows 10 force ppl to update it because it connect AMD software with it services they reason you need to use old AMD drivers because you did not update you windows . I bet you have lot of lags and FPS drop at this moment or barely getting 60 to 70 FPS.
also if you turn off you update without using the software that I use windows will update itself if you like it or not.
FOR GOD SHAKE AMD FIX YOUR DRIVERS.
It’s pointless to try establishing a new era for your you power and having them crippled by your drivers.
I got a RX5700. Normally runs the Rainbow Six Siege at 2k/230fps in ultra graphics settings. BUT especially since the update, the game crashes during the match and a window pops up indicating that ‘there is a driver timeout’. Following the obvious steps, i found out from windows reporting error system, that Radeonsoftware.exe was crashing unexpectedly creating all these game crashes.
Of course i DDU and reinstalled the latest recommended adrenalin like a million times.
Of course i tried everything in the nature of reinstalling and installing back.
And i come back to my original question:
Why can’t you at least, support your own GPU? I’ve always disliked NVIDIA, but at least they keep their costumers happy, long after their purchase. Please don’t make me choose something i don’t want.
P.S. I hope to fix all the issues with RX5700, so one day i can download and use the drivers without always having to tweak something…..
Never had a single problem with my older Nvidia Asus GTX570 DirectCu II (Massive 3 slot cooler), it did around 25% OC over stock clocks with 1.2 Vbios flashed at 60% fan which isn't even much noisy, was hitting just around 80Cº max even with the massive OC and was giving me above 1.8 Tflops instead of the stock 1.4.
When I went for AMD all the problems started including black screen not even being able to go to the bios as UEFI wasn't Working, had to put back the Nvidia to change bios to Legacy...
Then back with the AMD, bad texture problems on grass and many other stuff even until now and a lot of ppl also have the same issues across multiple different AMD generations and was never in the known problem list even with many reports.
And the worse was really that I only had to wait almost 2 years to be able to use it because drivers were broken and nothing would play since was all Crash to Desktop, BSOD, Green screen of death, Red screen of death... Even sent it to RMA for Asus and came the same GPU saying it had no problems and they stole my box and accessories since the store sent the original case and should have been a cardboard one even though that still very bad practise from Asus.
AMD sadly broke it again for like the hundred time already while I never had a single problem with the Nvidia for many years even at Huge OC and maximum voltage with custom bios to allow more voltage, it's still working...