ASUS TUF Radeon RX 5700 XT Crashing

[AfonsoJR]

Greetings people!

 

I know I know, the ASUS TUF RX 5700 XT has huge problems, I know, but at the time 400€ + 45€ cashback looked so much of a deal that I didn't even research the card

 

Fast forward to today, and I've RMA'd the card once (and was one whole month without a decent GPU), but luckily I still had my laptop to save me on some needed work.

The result of the RMA was none other than another card that was exactly the same, but.... I had some hope left in me! 

 

I installed the GPU and low and behold, crashing again! I don't know if you're familiar, but it is a crashing that occurs after intense gaming, or, intense something (even folding crashes my computer).

 

I have two monitors, one DP (turns black) and one HDMI (turns green) and the audio does that "stuck" sound.

 

So far I am in the process of starting another RMA, I will try at all costs to get rid of this RX 5700 XT, but with the whole pandemic situation, things will definitely take longer...

 

Which is why I come here, in search of a solution, or a glimpse of one.

 

I even took advantage of the situation to perform a clean windows install with the latest drivers (20.2.2), but I was met with the same fate of always...

[AfonsoJR]

@j0hn0b @LIGISTX After my rant on the folding thread I've decided to create this one.

 

@LIGISTX If this is a bad card, it is already the second one... Regarding PSU, I doubt it, since I own a Corsair 650W Gold PSU and I can hardly imagine my system ever requiring that much... right? Either way, they gave me a new card after I sent my first one back, so shouldn't be that.

 

@j0hn0b The crashes are better described above and I've tried running everything stock, overclocked, underclocked, undervolted, everything, but it all crashes

 

Is it just a bad card, and a second one at that? This was my first AMD experience, and so far it has been far from great...

[papajo]

try this 

 

what you also can do is improve the airflow and cooling of your system in general 

 

This card has not a good cooling solution so if you are goignt to stress it 24/7 with foldit and then when not folding then gaming... you asking too much from this card's anemic cooling...

 

So if in addition to that you have a dusty box in a corner somewhere with limited airflow and its hot where you live then i have bad news for you... try to improve your cooling situation in general or stop folding.

[xAcid9]

What is your card thermals? Junction,Mem, VRM, etc. 

Also what PSU?

[AfonsoJR]

5 minutes ago, papajo said:

try this 

 

I was afraid of this solution... So my only shot is voiding my warranty... In that case I think I'll just suffer through this and wait for the RMA process, since right now it is my best choice.

 

5 minutes ago, xAcid9 said:

What is your card thermals? Junction,Mem, VRM, etc. 

Also what PSU?

PSU is a Corsair 650W Gold.

 

No idea of Mem temps (how can I check them?), but junction temps can reach 95ºC last time I checked. 

[papajo]

7 minutes ago, AfonsoJR said:

I was afraid of this solution... So my only shot is voiding my warranty... In that case I think I'll just suffer through this and wait for the RMA process, since right now it is my best choice.

a) I editet my initial post take a second look on what else to do

 

b) you wont void your warranty by removing the cooling system to add some pads for the ram. 

[xAcid9]

8 minutes ago, AfonsoJR said:

PSU is a Corsair 650W Gold.

 

No idea of Mem temps (how can I check them?), but junction temps can reach 95ºC last time I checked. 

I think GPU-Z can show memory temp unless Asus remove or hide the sensor. 

 

95c for junction temp is still ok. What kinda crash you experiencing? Hard lock? Software crash?

[AfonsoJR]

10 hours ago, xAcid9 said:

I think GPU-Z can show memory temp unless Asus remove or hide the sensor. 

 

95c for junction temp is still ok. What kinda crash you experiencing? Hard lock? Software crash?

It is a hard lock and rebooting is the only option. 

[BTGbullseye]

28 minutes ago, AfonsoJR said:

It is a hard lock and rebooting is the only option. 

Try turning down the power limit in the Radeon software. It's under Performance > Tuning, and it's in the lower right corner. Set it to -25% and see what that does for ya. (you can also set a slightly more aggressive fan curve if you want)

 

Another thing is, what games are causing this problem? Anything using a Ubisoft originating game engine is inherently unstable on AMD hardware. (it's a Ubisoft problem that they refuse to fix) You can only really fix that problem by reducing the texture and/or shadow settings, while also limiting the performance of your GPU.

[AfonsoJR]

50 minutes ago, BTGbullseye said:

Try turning down the power limit in the Radeon software. It's under Performance > Tuning, and it's in the lower right corner. Set it to -25% and see what that does for ya. (you can also set a slightly more aggressive fan curve if you want)

 

Another thing is, what games are causing this problem? Anything using a Ubisoft originating game engine is inherently unstable on AMD hardware. (it's a Ubisoft problem that they refuse to fix) You can only really fix that problem by reducing the texture and/or shadow settings, while also limiting the performance of your GPU.

 

Hmm, -25%? That's agressive, but I'll give it a go and give feedback!

[BTGbullseye]

27 minutes ago, AfonsoJR said:

 

Hmm, -25%? That's agressive, but I'll give it a go and give feedback!

It shouldn't impact performance by more than 5%, but can usually increase stability massively.

[AfonsoJR]

47 minutes ago, BTGbullseye said:

It shouldn't impact performance by more than 5%, but can usually increase stability massively.

Sooo, I tested this, and after noticing the expected high memory temps, due to this ASUS model, I also added an aggressive fan curve, but I eventually crashed either way...

[AfonsoJR]

I was logging everything with GPU-Z and got these pretty normal values on crash time:

 

        Date        	    , GPU Clock [MHz] , Memory Clock [MHz] , UVD Clock [MHz] , UVD Clock [MHz] , GPU Temperature [°C] , GPU Temperature (Hot Spot) [°C] , Memory Temperature [°C] , GPU VRM Temperature [°C] , Mem1 VRM Temperature [°C] , Mem2 VRM Temperature [°C] , Fan Speed (%) [%] , Fan Speed (RPM) [RPM] , GPU Load [%] , Memory Controller Load [%] , Memory Used (Dedicated) [MB] , Memory Used (Dynamic) [MB] , GPU only Power Draw [W] , GPU Voltage [V] , Memory Voltage [V] , CPU Temperature [°C] , System Memory Used [MB] ,
        2020-03-22 16:15:07 ,        1983.0   ,           1750.0   ,           0.0   ,           0.0   ,               59.0   ,                          70.0   ,                  74.0   ,                   56.0   ,                    47.0   ,                    48.0   ,              55   ,                2535   ,         92   ,                       17   ,                       1095   ,                       73   ,                 136.0   ,         1.075   ,            0.850   ,               42.0   ,                  9642   ,

 

[BTGbullseye]

1 hour ago, AfonsoJR said:

I was logging everything with GPU-Z and got these pretty normal values on crash time:

 

        Date        	    , GPU Clock [MHz] , Memory Clock [MHz] , UVD Clock [MHz] , UVD Clock [MHz] , GPU Temperature [°C] , GPU Temperature (Hot Spot) [°C] , Memory Temperature [°C] , GPU VRM Temperature [°C] , Mem1 VRM Temperature [°C] , Mem2 VRM Temperature [°C] , Fan Speed (%) [%] , Fan Speed (RPM) [RPM] , GPU Load [%] , Memory Controller Load [%] , Memory Used (Dedicated) [MB] , Memory Used (Dynamic) [MB] , GPU only Power Draw [W] , GPU Voltage [V] , Memory Voltage [V] , CPU Temperature [°C] , System Memory Used [MB] ,
        2020-03-22 16:15:07 ,        1983.0   ,           1750.0   ,           0.0   ,           0.0   ,               59.0   ,                          70.0   ,                  74.0   ,                   56.0   ,                    47.0   ,                    48.0   ,              55   ,                2535   ,         92   ,                       17   ,                       1095   ,                       73   ,                 136.0   ,         1.075   ,            0.850   ,               42.0   ,                  9642   ,

 

What software/game were you using at the time?

[papajo]

2 hours ago, AfonsoJR said:

I was logging everything with GPU-Z and got these pretty normal values on crash time:

 

        Date        	    , GPU Clock [MHz] , Memory Clock [MHz] , UVD Clock [MHz] , UVD Clock [MHz] , GPU Temperature [°C] , GPU Temperature (Hot Spot) [°C] , Memory Temperature [°C] , GPU VRM Temperature [°C] , Mem1 VRM Temperature [°C] , Mem2 VRM Temperature [°C] , Fan Speed (%) [%] , Fan Speed (RPM) [RPM] , GPU Load [%] , Memory Controller Load [%] , Memory Used (Dedicated) [MB] , Memory Used (Dynamic) [MB] , GPU only Power Draw [W] , GPU Voltage [V] , Memory Voltage [V] , CPU Temperature [°C] , System Memory Used [MB] ,
        2020-03-22 16:15:07 ,        1983.0   ,           1750.0   ,           0.0   ,           0.0   ,               59.0   ,                          70.0   ,                  74.0   ,                   56.0   ,                    47.0   ,                    48.0   ,              55   ,                2535   ,         92   ,                       17   ,                       1095   ,                       73   ,                 136.0   ,         1.075   ,            0.850   ,               42.0   ,                  9642   ,

 

well these particular temps are actually fine... are you sure your crushes are because of temps ?

[BTGbullseye]

1 hour ago, papajo said:

well these particular temps are actually fine... are you sure your crushes are because of temps ?

He did say they looked normal, and didn't say he thought the temps were the cause there.

[GreyestGoat]

Wow what a POS, this is something i would not expect from ASUS. Are you able to send it back?

[BTGbullseye]

32 minutes ago, GreyestGoat said:

Wow what a POS

It could just be that the program he's using is unstable... Plenty of games like that. It's not a "POS".

32 minutes ago, GreyestGoat said:

this is something i would not expect from ASUS.

I would. They aren't really any better than any of the rest of the AIBs for these things.

32 minutes ago, GreyestGoat said:

Are you able to send it back?

He already RMAd it once without any change, so now he's doing the logical thing and trying to figure out what part of the rest of his system might be causing the problem.

[papajo]

16 hours ago, BTGbullseye said:

He did say they looked normal, and didn't say he thought the temps were the cause there.

yea I saw that later on but was too bored to edit the post at that time hahahahha

[JeannieFawkes]

I had a looooot of issues with a 5700XT but finally got it stable. I had black screen while playing. Have you tried the driver 19.12.1? It is what fixed my problems... Updating it only made it worse in my case...

[BTGbullseye]

18 hours ago, JeannieFawkes said:

I had a looooot of issues with a 5700XT but finally got it stable. I had black screen while playing. Have you tried the driver 19.12.1? It is what fixed my problems... Updating it only made it worse in my case...

Also, the Logitech Ghub is extremely bad, and will crash the 5700 XT's drivers. I just figured this out a couple days ago when it wasn't even letting me game for 3 minutes.

 

Went from about 15 minutes before crashing on The Division 2 when severely underclocked, to being able to run it well for over an hour with a borderline unstable overclock. (the only difference was Ghub running or not)