General Discussions

[CorfFerno]

So I hope I've posted this in the right forum (apologies if not).

By the looks of it (via google, reddit etc) I'm not the only person who has issues with 'Wattman' crashes. But I'm generally only getting them in one game: Total War: Warhammer 2 (TW:W2)

My system specs are:

OS: Windows 10
CPU: AMD Ryzen 5
RAM: 16.0GB
Motherboard: Gigabyte B450M DS3H-CF (AM4)
Graphics card: AMD Radeon RX 5700 XT
Power: Novatech 750W ATX Power Supply

My GPU is currently sitting on about 57c temperature but, when playing TW:W2, generally goes up to about 90c. I've recently bought an EVGA CLC 120 Liquid/Water CPU Cooler that I'll be installing when I get chance to Try and bring the temperature down a bit.

But I'm still getting these Wattman crashes and can't figure out why.

I've used the AMD cleanup utility app to remove previous drivers and reinstalled them again. But I'm no closer to cracking it.

The Windows Reliability History isn't showing anything particularly useful either. Or at least I may not totally be understanding what I'm seeing in there.

Does anyone have any advice for this or tips of what I should be looking for?

Much appreciated if so!

[iTruji]

@RPX100 

Before starting, on my fresh windows 10, to install any adrenaline software, two days ago, for first time, I installed only the drivers, cause I read that some people fix problems. I use afterburner to set the undervolt, underclock and fan curve. Since this, I haven't any freeze.

I'm not trying custom profiles for games before, but I see that adrenaline doesn't like ours 5700 xt. For example, my system can freeze instantly (black screen and must to power off and reboot) when I turn on the adrenaline metrics playing a game (not always).

I asked you for xbox bar cause I used adrenaline to capture video from the games and xbox bar has a similar feature. I'm using this option and works fine.

Sorry, but I don't undestaint what's the problem with light games that downclock the GPU. Is downclock important  for stability? And about ULPS, what's the impact with stability? When I did this changes in settings in the past, I didn't undestain what I was fixing (my fault).

[RPX100]

@iTruji Radeon Software really does not like the RX 5700 XT. I agree. (-.-)'

I am also -sometimes- using the Xbox Game Bar to capture videos.
It works better than Radeon Software for me too.

---

ULPS = Ultra Low Power State
Which is one of many power saving features of your graphics card (on a driver level).

My graphics card is rated at 1905 MHz boost clock, by manufacturer.
But some games that only require "light GPU load" only use like 800 MHz GPU core clock.
This can lead to "micro stutters" or lag in some cases, when you turn the camera or switch areas.

This becomes even worse in demanding games like No Mans Sky, where the GPU load
may vary between 99% and 30% depending on the camera view and area you are in.
In order to prevent these issues, I am using custom profiles that force my clock to stay around 1905 MHz,
instead of allowing the energy saving feature to downclock my card to 800-1000 MHz.

So in short: ULPS can cause stutter/lag in some cases.
But it would a useful feature to save energy, if it would work correctly.

---

not sure about the impact of ULPS on stability. I can not disable this feature on my own card,
and I do not have any freezes currently... so I guess it works well for my card,
if you ignore the fact, that I have to manually lock my boost clocks to have stable performance.

[iTruji]

@RPX100 

Thanks for the ULPS and downclock explain!!

How do you lock the GPU clock? Setting the same mhz (1905) on every state?

[RPX100]

@iTruji setting the same MHz for all 3 is a bad idea and is causing instant system freeze.
Seems to be a driver limitation, because of the power saving features.

The correct way, that is working fine with my card:
1880 MHz - 1025 mV (P1) = default clock
1905 MHz - 1040 mV (P2) = game clock = actual clock under load
1950 MHz - 1055 mV (P3) = boost clock

Using these settings, results in game clock of 1900-1910 MHz for me.
Only when the game is at less than 30% GPU usage, it drops to 1875-1885 MHz.

[iTruji]

@RPX100

Thanks again, noted!!

[colesdav]

You should just send the crap GPU back to the manufacturer and NOT put up with it anymore. 

AMD are making massive profits because people like you (and me in the past) put up with bad AIB GPU and bad unstable drivers. 
I cannot even use 20.2.1 driver on my PC at all. 
I just cannot be bothered with AMD now.

[mackbolan777]

Well it seems as though the very same settings to get stable I posted in a prior thread work. Moving all the clocks closer together within 100Mhz and you really only need to set one voltage, since the others fall in line after you set the max. I also made the point about the RX 5700 XT's top clock staying under the 2000Mhz mark since it usually becomes unstable or actually loses performance above 1950mhz. Like you report your in game clock never comes close to the 1950 max as per GPU BIOS limit which is set around the 1920Mhz area. Another secret AMD doesn't mention is that VRAM needs to set 12Mhz higher than whatever you set it at. Like 1750=1762 would be more stable.

This moving the clocks closer method also works with the newer 6000 cards as well. It stops the card from spiking clocks and causing sudden voltage surges that trip the PSU's safety measures. It is unique to see a 5000 card running a driver newer than September without crashing or other issues. It also nice to know that AMD still insists on using 3 clocks on the 5000 series cards rather than 2 like the 6000 series do.

Why have "power saving features" on an "enthusiast" card? It's like buying a Corvette with a governor, or expecting it to get 40 MPG. Really confusing. Even with my somewhat aggressive OC on my RX 6800 I'm pulling 32W idle vs. 30W stock. Same as my former RX 5600 XT. So where's this power savings at anyway? 

No offense intended RPX, I just find it funny how we all seem to post the same solution 1000 times on here and it comes off as "fresh" every time. Lol.

[dekho]

Getting the same error for some months already and trying to fix it by myself reading stuff. I haven't made any improvement untill I found this thread.

I'm going to post my current options and pc components and hopefully you could tell me where to modify starting from this , since the same path might not work for me.

Also going to provide you with some info to understand my frustration more. This is desktop is supposed to be used for extreme gaming but so far I was just able to enjoy mediocre games at mediocre settings. Hopefully you will try and help me. I have this error since I've got the pc around ( on 27 april 2020 to present). 

Currently I have the version 20.11.2 , previously tried to update it to latest one as well (21.2.1) but that one was the most hellish until now as I would have got around 6-7 restarts a day from the mediocre games that used to work fine as dota 2 and cs go for example. So I uninstalled everything and then installed 20.11.2 ( if you know a tip for me to choose easier which version I want to run, let me know please , I would like to try the one specified earlier 20.8.3 ?)

These are my components:

AMD Ryzen 5 3600 3.6GHz, 6 nuclei, 16GB DDR4, 1TB SSD, RX 5700 XT 8GB GDDR6

Recommended for extreme gaming

Processor:

Producer: AMD

Processor: 3600

Socket: AM4

Nucleus: Matisse

Family: Ryzen 5

Nr. Nucleus: 6

Nr. Threads: 12

Frequency: 3.6 GHz

Frequency turbo max ( not that I ever plan on using that 4.2 GHz

Manufacturing technology: 7nm

Video card:

Producer: ASUS

Video card : Dedicated

Family: Radeon

Model: RX 5700 XT

Memory type: GDDR6

Memory size: 8GB

BUS memory: 256

Motherboard:

Producer: ASUS

Producer chipset: AMD

Chipset: B450

Nr. SATA-III: 6

Slot M.2: 1

Memory:

RAM type: DDR4

RAM capacity: 16GB

Frequency: 3000 MHz

Slot 1: 8GB

Slot 2: 8GB

RAM memory support:

Memory slots: x4

Empty memory slots: x2

Maximum memory: 64GB

Storage:

SSD

Capacity SSD: 1TB

Type SSD: M.2

Interface SSD: PCI Express 3.0 x 4

Carcase:

Producer: Gamemax

Model: Abyss TR

Type: Middle Tower

HDD storage unit support: 2

Suport unitati de stocare SSD: 4

Source:

Souce: Dedicated

Power source: 650 Watt

Efficiency: 85%

Certification: 80+ Bronze

Cooling system:

Type: SE-224-RGB

Air

*Things I've changed since reading this thread*

I really like this setup especially for fans as it stay around 38 c when idle, 40 - 41 c when I'm in a mini game, around 46 c when I'm in cs go and around 48-55 when I'm in dota 2 which is very nice as it was going to even 70 with the old preset , and in games like Modern Warfare Warzone would go to 80-ish 90-ish and restart due to the error. I tried apex before this options and it was going pretty hot 90 and crash (couldn't play with my friends). And at the start in april I've tried to play The witcher 3 and

crashed in matter of 10 minutes on the lowest graphic settings.

After I tried these setups on performance tunning the temperature is great. For example today I tried some APEX (I don't plan on playing this too much but this is something that my pc should be able to run it without issues so I'm just testing around). If this will work fine,  Warzone will too)

Today I tried to enter APEX with these settings from the pictures above and the temp was just great holding at around 50 mb 55 didn't stayed to check and it started at around 44-46. At some point very early in the game , I think I played only 2 minutes the pc crashed. 

So the temperature wouldn't be the problem here I would say.

This is everything to give you an idea of what I have. I'm looking for a solution for the error with the current stuff I have, I don't want to buy or change anything, if possible please help me, this error is eating me alive.

Please if you could provide me with some steps starting from here I would be FOREVER THANKFUL as this is something that upsets me a lot.

[mackbolan777]

Ok, so download the 20.8.3 driver and save it from here: AMD Radeon™ RX 5700 XT Previous Drivers | AMD

Download DDU from here: Display Driver Uninstaller Download version 18.0.3.6 (guru3d.com)

Disconnect from the internet for now.

Install DUU and reboot into safe mode by holding the "shift" key while selecting "restart". At the blue screen, choose "advanced", "startup options", the PC will reboot to another blue screen select "safe mode". Install DDU and select to uninstall the GPU driver. Leave the other options alone and remain offline.

The PC will reboot, now install the 20.8.3 driver, remaining offline. The PC will reboot, remain offline. Go into the Radeon settings and disable the auto update feature, preventing the software from automatically installing newer drivers. You can always read release notes for newer drivers to see if there are any fixes or anything new for your card, there won't be because it is EOL. So if this driver works well, stick with it.

Ok, now for those settings. You have an RX 5700 XT so your clocks/voltage need to be a tad higher. Set your max clock to 2000, mid clock to 1800, low clock to 1700, set voltage to 1.107mv, add 12Mhz tot he VRAM. Try it out with everything else you show in the photos set they way you have them, looks good. 

[iTruji]

@mackbolan777 @RPX100 

Yesterday I tried this settings but Valorant crash (furmark doesn’t). Maybe lower mV?

[mackbolan777]

Your voltage is too low for a RX 5700 XT, try going to 1.107mv for max voltage, the other voltages will fall in line.. The rest looks ok.

[iTruji]

@mackbolan777 

Would I set P3 to 2000 instead 1905?

[mackbolan777]

Pretty much what @RPX100 said. You can try 2000, if yours is a Strix it's rated to 2035 but it will run better at 2000 or less. Typically 1950 is the ceiling and these cards actually respond worse to more voltage after a point. If your crashing still at 1.107mv, go up +.10mv until stable. So next would be 1.117mv, 1.127,1.137, etc. you'll stop before any more most likely. The +12Mhz to the VRAM is an undocumented AMD secret I stumbled on recently. The VRAM needs +12Mhz to whatever setting you try to apply for stability.

[colesdav]

It's all Voodoo until AMD fix their drivers, and given their track record, they won't.

[iTruji]

@RPX100 

@mackbolan777 

My GPU is a MSI gaming X with 2100 "stock". For the balance between fps and temps, I find that 1905 is my choice. I'm adding more mV following your recommendations.

This week my system is more stable than before with 20.11.2 (only the drivers, no adrenaline, but using msi afterburner to undervolt/underclock), with a few crashes in Valorant. I tried before with 20.8.3 but had more crashes and don't know why...

In afterburner, I can set the three P states (for exemple 1905, 1805, 1705) but saving and applying this profile, the gpu clock goes down 1700 mHZ (it seems that I'm missing a step). With adrenaline I can do the same and works fine, but I'm testing without adrenaline at this moment. (@RPX100 can you help me with the AB settings?)

Using AB, when I set the VRAM to 1765, my screen starts to flickering (idle and gaming). Undo to 1750 and flickering stops (strange). I am tempted to use the latest version 21.2.1 that seems to fix flicker when AB is running.

I'm registering all my crashes in valorant:

- Always crash when the round starts (or few seconds later) when i'm pressing the "w" key and agent start to run. Sometimes is when I'm entering/exiting into a tunnel or closer/open scenario (when there are a lot of changes in fps and gpu clock).

- I didn't limit fps in game but 60 fps in menu ( I have a 27 2k 170hz monitor). Seems that limiting fps in game to 240 (usually go 300 fps and more without limits) made my system more stable.

Thinking about this, I'm pretty sure that the "stock" P states and wrong fps settings in Valorant causes important voltage/watts peaks, and maybe the crashes.

This week I will try with the modified P states (if i can set it correctly with AB).

[kalckamlo]

Thanks for answering.

Nox   Vidmate   VLC

[iTruji]

@kalckamlo 

Sorry, are you asked to me?

I'm on the latest win10 pro version (20H2).

[RPX100]

@iTruji depends on the type of crash, that you are getting:

crash type #1: APPCRASH to desktop:
- app freezes, sometimes followed by a short black screen or screen flicker
- display driver resets itself and you are back on desktop
-> solution: increase voltage +10 mV (repeat until crash is gone)

crash type #2: black screen and system freeze with system reboot
- app freezes and system freezes and screen stays black
- system will either restart itself or you have to reboot yourself
-> solution #1: decrease core frequency
-> solution #2: decrease memory frequency

Instability on the Radeon 5000 series can be caused by:
- GPU core clock too high
- GPU mem clock too high
- GPU core voltage too low (or clock too high for that voltage)

In your case, judging by the screenshot: increase voltage a bit.
Leave your P3 at 1905 for now. Just increase voltage.

[dekho]

Hello again,

I followed your resolve step by step as I just finished I'm typing this message to you.

*this is how it looks after*

Something I would have to add is that I tried to " Set your max clock to 2000, mid clock to 1800, low clock to 1700, set voltage to 1.107mv, add 12Mhz tot he VRAM. Try it out with everything else you show in the photos set they way you have them, looks good. " , as I tried this the highest clock from far right I could edit , the middle one its greyed out I couldn't change nothing at it, so I jumped to the first one and placed the frequency at 1700 as you've said. This made the 2nd clock change at a higher value and then I modified the voltage from the far right from 960 to the one recommended at 1.107 and then went to VRAM and added +12, so from 1800 to 1812. The next following 5 or 10 seconds it crashed but without reset, it just shutted off the radeon software and automatically placed it on default. 

Currently I'm on the same settings like on the screenshot above. Any tips for that ? Or I should give it a try in games and see if I get it again ?

[dekho]

Just gave it a try with the options from the screenshot above and the driver you've recommended me.

I was able to play apex starting from Low setting to medium and then high. Tested little by little to check out if it works so far no error the temperature never passed 58. All good.

Went into CoD Warzone on Low played a game. No worries it worked. Then switched everything to high everything , still was able to play, no errors just joyful gameplay. Temperature never passed 58.

Been able to play my mediocre games csgo , dota 2 , world of tanks with best options. No problem so far. I'm really happy about the outcome , I hope it lasts forever. I will come back if not !

THANK YOU FOR HELP

[mackbolan777]

You need to open the fine tuning area if you didn't already. Most not all RX 5700 XT's will handle 1800 VRAM, skip the +12 if it causes an issue. However, even at the stock 1750 you should have zero problems. My RX 5600 XT handled 1800Mhz VRAM without any issue.

You should be able to type the numbers for your clocks in manually, if not it might be a BIOS limitation for your brand card. Another OP mentioned using 1905 as the top clock, I would go even numbers like 1900/1800/1700. The lowest clock may or may not matter since that should be the 2D clock for non-gaming. So the middle and top are most concerning to get right.

If you're still crashing, leave the clock settings as is and allow the voltage to go back to stock, see if your stable. It's not really "Voodoo" as one put it, it's too many "partner" cards out with different PCB layouts and IC's. So an MSI will be different than an XFX or a Gigabyte, etc. It boils down to researching the build quality, reading reviews or guides on other forums to get the answers one needs. No "one size fits all", just a "one size fits most".

[dekho]

I'm trying it like you say M. (photo below)

[dekho]

Let me know if voltage should be lower or higher, if its not working I guess I will have to make arrangements to send it back. Also thank you for helping me guys. Means a lot to me.

[mackbolan777]

How to OVERCLOCK and UNDERVOLT RX 5700XT | ADRENALIN 2020 Easy Guide, Tutorial - YouTube

This video is the best OC guide online for most RX 5700 XT cards. It's the basic source for nearly all the advice we gave you.

Try a different PSU or swap the card into another PC to test it.

You can try installing the driver, no Adrenaline, by letting the driver install and then copying the install location and canceling the rest. Then going into device manager and right clicking the GPU and selecting "update driver", then pasting or typing the driver install location into the box manually.

After that it seems like an RMA might solve it. If your card doesn't work as good as the one in that video, you got problems. He even runs it on the loose side as far as VRAM and whatnot. He allows for the fact not all cards will achieve his OC and tells you alternate settings in a range.

[iTruji]

Hi!

It's seems that I've fixed my GPU.

I visit igorslab forum and open a post. I guy offered to "cook" a VBIOS for my MSI Gaming X with some settings to made the card more efficient and stable.

The most important fix is for the vRAM: he was applied Apple Inc. vram timing straps (from the macs vbios).

With the "update" VBIOS,  no more crashes, black screens,... gaming or working, regardless the amd driver version or the settings in adrenaline. I'm on 21.3.1 and test with standand or gaming profile and no problems playing Valorant/Warzone,...

It's awesome, now I can enjoy my GPU after 5 months... 

[mackbolan777]

Surprising a trustworthy person made a vBIOS custom for you. That's a Russian site but has good info. So, how are the Mac timing straps? Does it lock your VRAM lower or the same but with better timings? Can the card still OC?

[iTruji]

The site is from Germany (not Russian), and has a lot of info and a good community.

I just trusted the person who wanted to help me and his vBios worked.

About the RAM timing, read this thread: https://www.igorslab.de/community/threads/gigabyte-5700-xt-bios-mod-fails.3304/post-86493

About the VRAM, he sent me 4 bios with vram between 1870-1950. I only test the first with 1800 (without touching settings in adrenaline) and max. 1870. Very stable.

Using adrenaline, it’s possible OC but I forgot to use adrenaline settings for a long time (defaults works awesome). 

[dekho]

Sadly to say. Its the best result so far but still happens after 4 hours or so of playing restarted again with wattman crash , also today while not playing just working on it crashed again. Do you have any other advice to try on settings ?

Kind of sad for still happening but atleast I get to play some matches between the errors

[RPX100]

@dekho remove the VRAM overclocking.

Set your VRAM back to the factory clocks (should be 1750 MHz).

[dekho]

Tried the VRAM at 1750 still crashed , 1812 still crashed.

[RPX100]

Are you using any of these Radeon Software features?
- Radeon Boost, Anti Lag, Enhanced Sync, Image Sharpening
- disable them, if you have them active.

Did you select gaming / e-sports / energy saver profile?
- "standard" profile works best for stability

Do you have any custom game profiles?
-> I have encountered crashes while using them
-> my card was only stable on "global" performance profile

---

You could also try and check if your issue is related to your system memory.
-> I think this only is important if you are using AMD processor. Ignore this for Intel.
If you have XMP enabled in BIOS, try to disable XMP and instead:
-> enter the correct frequency and voltage yourself

[dekho]

I'll give it a try

[dekho]

Tried those RPX worked for approx 1h and then it crashed twice in 15 minutes 

[mackbolan777]

@iTruji 

The settings inn that fix apply to a RX 5600 XT, so your voltage would need to be higher. 1.050mv sounds about right for a RX 5700 XT. When you go removing drivers, be sure to use DDU each time. You should only need the 20.8.3 or 20.9.2 as the newer one's are geared towards the 6000 series and offer no gains to your card, if not more headaches. Yes, you will lose a few FPS and that's more or less by design to lower the temps and increase stability but 950mv might be too low.

The video on Ancient Gameplays via YouTube explains the process in the best detail. 2100Mhz is a bit crazy for a factory OC and probably a "wish list" item from MSI. It's unrealistic with the cooling solution they or any manufacturer uses. In the video I describe above, the OP says to lower the clock to ~1900Mhz then do the undervolt of ~1.050mv or so. Fan curve can help with the heat, try to go more aggressive with less jumps per degree, this lowers the polling by the controller and increases stability.

 @RPX100  makes a point with VRAM, don't mess with it or try his setting for that, the 5000 series aren't good with VRAM OC at all. They have plenty of VRAM speed at ~1750-1800, which translates to 14.5Gbps-16Gbps.

Make each change slow and don't flip driver to driver like a madman. If you had a max stability at 1905/950mv that's what I'd use, the few FPS aren't worth the aggravation. But if you still had stability at 1999/1.050v, that's ok too. Mess with the fan curve and you should be good to go with that. The factory 2168/1168mv is insane, however even at the high 95c, still within operating range. A fan curve adjustment could lower that temp to 80c and be stable. Remember to always use a game as your "real world" stability test, because these 3D Mark TimeSpy type tests are the worst case scenario and your card isn't going to be running at those extremes as long or ever.

Games like Valorant are very CPU demanding, so if you're crashing out there, it could be a power supply issue or CPU cooling/setting issue you haven't looked at. You should always use 2 X PCIe power plugs vs. the split dongle for the GPU. One cable can only supply 300w and 9A. The card can pull the wattage relatively safely but you're looking at a ~14A draw which exceeds a single cable's capacity. Cards of these class pull roughly 2.5W per frame, so let that be a guide as why 2 are better than one cable wise.

[iTruji]

@mackbolan777 

Maybe my problems is to jump driver version to another version too fast... 

I will start with 20.9.2 and test the undervolt and maybe the underclock later.

With only the GPU undervolted (2100 mhz and 1116 mV) the 95ºC is with the fan at 100% playing COD Warzone and higher than 100ºC using furmark. In Valorant, temps are down 80ºC.

I try again with 1905/950 and going slowly with the test. But if I have a crash, what would I do? Up +10mV? I tested in furmark and it's really stable for 15 minuts but as @RPX100 says, testing in real game can crash even if in furmark is stable.

I will not touch the VRAM clock.

My PSU is a corsair 750 gold and the gpu has two different wires and not the split. Despite this, I'm using this cable a month ago: https://www.amazon.es/LINKUP-Juego-extensión-Cable-Personalizado/dp/B07J9Q9RMG/ref=sr_1_1_sspa?__mk...

And CPU (i5 10600k without OC), has a Noctua D15S and temps are low.

[Antreas071]

Your card runs stably 2000mhz max clock with only 1025mv? When i set it below 1080mv @2000mhz games crash and when i set vram over 1780mhz i get blackscreens. I was thinking of watercooling it but when i saw what you got, i dont think that it deserves to be watercooled.

[RPX100]

@Antreas071 hello there!

I have a 24/7 profile with low energy consumption that I play most games with:
943 mV and 1840 MHz (results in a stable 1790 MHz game clock when above 90% load)

I also have a OC profile for "medieval dynasty", which can be very GPU hungry:
1030 mV and 1975 MHz (results in a stable 1925 MHz game clock when above 90% load)

Anything above 1925 MHz game clock was causing crashes. Even with 1200 mV!

manufacturer specs for my card:
Boost Clock: Up to 1905 MHz
Game Clock: Up to 1795 MHz
Base Clock: 1580 - 1650 MHz

And I am able to achieve those specs, when I focus on the gaming clock.

I am not able to overlock memory on my card. Anything above 1750 MHz will cause frequent crashes.
Also please note that GPU hot spot temp should stay below 90°C.
And temp delta between GPU core and hot spot should stay below 15°C (less delta is better).

[iTruji]

Thanks @RPX100, for my 20.11.x drivers not works.

I’m trying with 20.9.1 and day two, perfect, except for a crash when i made an instant record (the first time). 

[mackbolan777]

All that anti-lag, radeon chill, stuff is of no help and can incur a performance hit or undesired operation. It's simply a "click n go" solution for those who want to experiment with features with a click rather than try each thing to see if you benefit. Persoanlly and Fabio would say the same, is to just use the "custom" and leave all the stuff off completely. I personally like the image sharpening at 20%, surface optimization has been on by default for ages, so I leave it on and no it seems to help with DX9 apps. Both have zero performance hit. The other options change color temp, turn things on like radeon boost or chill, things that may or may not affect performance but will affect visuals. Radeon Boost is really poor because it degrades the resolution purposely in an attempt to maintain frame rate, so image quality suffers to a point where I can see it and also causes micro stutters, from my testing. Anti-lag is useless period. Your ping has more to do with lag and possibly incorrect monitor settings or not having FreeSync on. I say 20.8.3 because that was the last driver that worked pretty much flawlessly with the 5000 series GPU's. It is noted that 20.9.2 is another possible one, it really depends setup to setup but 20.8.3 is proven as "tested" for that "fix". 

If your card is working as you like, leave it alone. I would suggest not using "profiles" in the "graphics" area to try to gain more than you get from using the manual fix. It winds up counter productive in the end. Your esports or whatever game will play/look the same or better than using a preset. Try it out if you want, just remember what you change so you can go back.

[iTruji]

Ok @mackbolan777 , good advice. 
I did an undervolt and decrease the gpu clock from 2100 to 2000, following the youtube ancient channel, and increase the vram from 1750 to 1800. 
I will try during the next week. 
Thanks!