RX 580 heat going up to 97

[Alan GE]

Hello !

 

I have this rx580 build .. i built it in 2018 .

 

So i recently had this issue when playing games (like warzone) the pc just shuts down immediately after loading the graphics.

 

I tried to stress test the card from AMD software and the temp went up to 97 and the power consumption went to 110+ w.

 

I'm adjusting the frequency so the pc doesn't shut down when i play games which is working .

 

I'm wondering what can cause this issue? is it the power supply ?

 

i have 650 cooler master mwe 80 bronze

 

the card is rx 580 strix oc 

 

 

 

 

 

[UnusualDevices]

1. Are gpu fans spinning? according to that screen 0 rpm but that may not be correct.

 

2. update bios/ddu/reinstall drivers

 

3. if that doesn't work you may need to repaste your gpu.

[Alan GE]

3 minutes ago, UnusualDevices said:

1. Are gpu fans spinning? according to that screen 0 rpm but that may not be correct.

 

2. update bios/ddu/reinstall drivers

 

3. if that doesn't work you may need to repaste your gpu.

odb mode is turned on i guess thats why its 0 rpm . 

the drivers /bios already updated . clean install with ddu.

[Dukesilver27-]

4 years, have you cleaned them within the 4 years you have it?

Take it to a repair center and ask them to replace the thermal paste, it is likely that the thermal paste has been used up.

After that, undervolt the card, I see that you don't apply undervolt, follow this guide.

[UnusualDevices]

it may be the psu, but that's also really hot for a 580.

[Alan GE]

10 minutes ago, Dukesilver27- said:

4 years, have you cleaned them within the 4 years you have it?

Take it to a repair center and ask them to replace the thermal paste, it is likely that the thermal paste has been used up.

After that, undervolt the card, I see that you don't apply undervolt, follow this guide.

I haven't cleaned it tbh ... it makes sense replacing the thermal paste .. i just tried to stress the card again it went to 92 in less than 5 seconds 

[Dukesilver27-]

6 minutes ago, Alan GE said:

I haven't cleaned it tbh ... it makes sense replacing the thermal paste .. i just tried to stress the card again it went to 92 in less than 5 seconds 

Yea, that card should not be that hot, it is highly unlikely that PSU caused it.

Temps should be between 80-90C, 97C is very close to max operating temps of most GPUs, probably the temps spikes to 100C+ so the GPU shut itself down to prevent damage.

And clock speed really affects power draw therefore heat output, therefore reducing max clock speed helped you.

 

In general, after 2 years (if you purchased it new), you have to clean it, if still toasty, then thermal paste replacement is in order.

While you're at it, clean the whole PC.

[BondiBlue]

5 minutes ago, Alan GE said:

I haven't cleaned it tbh ... it makes sense replacing the thermal paste .. i just tried to stress the card again it went to 92 in less than 5 seconds 

That's not normal at all. That's worse than a 5 year old RX 570 I own, and that card still has the original thermal paste and two failing fans. It's also sitting directly above a second GPU running at full power, so it's not running in ideal conditions. 

 

You should remove the cooler and reapply the thermal paste. Give it a good cleaning while you've got it apart. It sounds like your heatsink isn't mounted properly for some reason. 

[aren332]

I have a gtx 645 which used to spike up to 97 degrees too, and I replaced the thermal paste, cleaned it and now it barely reaches 85 degrees. After 4 years of use, clean your gpu cooler with a duster or a vacumm. Make sure you hold the fans so they dont overspin while cleaning it. And when you repaste the gpu, make sure that you use rubbing alchochol to clean the gpu itself, and put fresh thermal paste. Dont use too little thermal paste, its better to use more. I recommend grizzly or arctic mx silver paste, but for me noctua thermal paste worked perfectly fine.

[Alan GE]

On 11/12/2022 at 4:31 AM, Dukesilver27- said:

Yea, that card should not be that hot, it is highly unlikely that PSU caused it.

Temps should be between 80-90C, 97C is very close to max operating temps of most GPUs, probably the temps spikes to 100C+ so the GPU shut itself down to prevent damage.

And clock speed really affects power draw therefore heat output, therefore reducing max clock speed helped you.

 

In general, after 2 years (if you purchased it new), you have to clean it, if still toasty, then thermal paste replacement is in order.

While you're at it, clean the whole PC.

After replacing the thermal paste now the temp does not exceed 60 when stressing and idle 29 and the profile is overclock ..thank you 

[UnusualDevices]

On 11/11/2022 at 8:08 PM, UnusualDevices said:

1. Are gpu fans spinning? according to that screen 0 rpm but that may not be correct.

 

2. update bios/ddu/reinstall drivers

 

3. if that doesn't work you may need to repaste your gpu.

 

57 minutes ago, Alan GE said:

After replacing the thermal paste now the temp does not exceed 60 when stressing and idle 29 and the profile is overclock ..thank you 

great job! don't forget to select the solution!.

 

oh you selected someone else even though I recommended repaste first.

[Alan GE]

53 minutes ago, UnusualDevices said:

 

great job! don't forget to select the solution!.

 

oh you selected someone else even though I recommended repaste first.

ops .. now its fixed