Is my GPU defective?

[AverageDude000]

Hi everyone,

About a month ago I bought the parts for this system and went to a professional for the PC assembly.
Ever since I started using it, I seem to be having problems related to the graphics card.
I started playing a game and had multiple random crashes as well as a bunch of stuttering.
The hotstop temps were reaching 95°C so I made a custom fan curve in Adrenalin (because for some reason the fans weren't spinning fast enough in high temps).
The fan curve solved the temp issue and now hotspots were getting to 80°C max.
The only other thing I touched in Adrenalin was raising the GPU power limit to max (15%) because I read that it just allows the GPU to draw more power when it needs it.
Started playing again but the problems didn't resolve and one of the crashes was serious enough that I had to force shutdown using the power button and when it rebooted the graphics card had been disabled in the device manager.
I enabled it and decided to use AMD Cleanup Utility to reinstall the drivers in safe mode hoping things will get better.
That night, just when I pressed "Shut down" in Windows to turn off the pc, another crash and reboot occurred and the Adrenalin settings had reset (as they did with every crash, that's why I knew it was the GPU again).
I wasn't even gaming that time, it literally crashed when I pressed "Shut down".
This time I decided to use DDU on the graphics drivers and even update the chipset drivers because I had installed them from the MSI site and they were older than the the ones on the AMD site.
I was planning on doing a clean reinstall of the game as well but encountered issues even before that.
I was watching a youtube video and it started randomly stuttering (video and audio).
I put the Adrenalin overlay on and noticed that when the stuttering happened, the GPU Utilization, Clock and Voltage all dropped to 0 during the stutters.
Also noticed today that the GPU Memory Clock is also stuck at zero after a few minutes of using the PC (not just during the stutters but it's been at 0 since I noticed it).
I'm using the Adrenalin software to see all this information, could it be that it's inaccurate?
All this and I'm just watching youtube on firefox.

Additional info:

• OS: Windows 11 Pro 23H2
• I updated the BIOS when I first got the PC about a month ago and turned on XMP without touching anything else (no PBO or anything)
• I added an extra intake fan to the front of the case (current combination is 2 intake (front) and 2 exhaust (back and up))

Where do I even start figuring this out?
Is one of the parts defective?
Do I need to DDU again maybe?
Maybe a clean reinstall of Windows will help? (although I really want to exhaust all other possible solutions before that)
Thank you in advance and please try to be specific when it comes to suggestions because my knowledge is pretty average.

PCPARTPICKER Screenshot:


YOU CAN SKIP THIS PART:
Sorry if this has been a bit rambly, I tried to include as much relevant info as possible.
Truth be told, this situation hasn't been great for my anxiety because it feels like I paid all this money and can't even watch a youtube video let alone play the games that help with anxiety.
I would genuinely appreciate all the help.
Thank you again.

Edited February 20, 2024 by AverageDude000
additional info

[SUPR3M3 OM3G4]

95°C does sound quite hot.

Couple of quick questions please 

 

1. Was the card new or 2nd hand?
2. What are the temps like on your other components like CPU?
3. What is your average room temperature there?
4. Did you open the card to replace the thermal compound?
5. What case do you currently have the Card and PC in and with what fan combination? 
6. Did you have a different GPU running on your current install of windows before installing this GPU?
7. Could you also maybe share an image of your PC currently and where you have it on your desk and how it's positioned when you run it?

 

Once I have these details then it might help us troubleshoot a bit more to find out what's wrong.
 

[AverageDude000]

30 minutes ago, SUPR3M3 OM3G4 said:

95°C does sound quite hot.

Couple of quick questions please 

 

1. Was the card new or 2nd hand?
2. What are the temps like on your other components like CPU?
3. What is your average room temperature there?
4. Did you open the card to replace the thermal compound?
5. What case do you currently have the Card and PC in and with what fan combination? 
6. Did you have a different GPU running on your current install of windows before installing this GPU?
7. Could you also maybe share an image of your PC currently and where you have it on your desk and how it's positioned when you run it?

 

Once I have these details then it might help us troubleshoot a bit more to find out what's wrong.
 

1. Every PC component is new.
2. CPU is 40°C idle and was getting to 55°C at most during gaming (I think)
3. Average room temp is at 22-25°C
4. I got the parts and immediately went to a professional for assembly because I couldn't build it myself.
Nothing was altered in any way as far as I'm aware.
5. The case is a "be quiet! Pure Base 500DX ATX Mid Tower" and all I did was add another intake fan to the front (current combination is 2 intake (front) and 2 exhaust (back and up))
6. The current PC parts are the ones I had from the beginning and I haven't changed any of them (the PC is a month old after all)
7. I cannot share a pic at the moment but I have the case on the floor and there's enough space around it (I think)

The 95°C hotspot temp was during gaming and before I added the custom fan curve.
I think Adrenalin was bugging out and in combination with the Zero RPM feature wasn't running the fans adequately.
After I put the custom fan curve on, the hotstop temps went to around 80°C max (during gaming), so overall I don't think temps are the issue.
Idle (or light load) hotspot temps for the GPU are around 40-45°C (50°C when watching youtube).


Ps: Btw, the Zero RPM feature definitely doesn't work right because it's supposed to start the fans at 50°C (as shown in the fan curve graph) but I've noticed the fans start around 58-60°C.

[Chapunna]

Update bios version, also remove bios cmos battery and then put it back, check if you plugged gpu power cable correctly, reinstall a fresh windows 10, disable secure boot and download way older gpu driver like 4 - 5 months old, dont forget download all drivers from motherboards manufacturer site

[Chapunna]

Also set lower resolution 1080p, and use only one ram stick 16gb in different slot, and turn off xmp

[SUPR3M3 OM3G4]

Hmmm I see what you mean, thank you for the clarification.

Have you tried adding a small bit of voltage to the GPU and downclocking it a bit.
That might help troubleshoot, because if it's working fine, then it seems the chip might actually be problematic or not able to reach it's standard clock speed. 

If it's still crashing then, then I'd say the only other option left would be to do a clean install of windows and installing the latest drivers.
If that fails, I'd say your GPU is potentially faulty.

But before that, would you mind me asking if it's only crashing on specific games?
Or is it crashing on all games?

For example, my PC is crashing a lot when I play one specific game because there is a known issue with the 14900K on that game and the devs are working to fix it.
So if it's only 1 specific game, then it might not be your GPU 
Hopefully we can figure this out for you.
 

[AverageDude000]

@Chapunna Hi thank you for answering.
Can you please be a bit more specific on how those things will help me?
I have a 1440p monitor, why would I run it at 1080p for example?
Also why use only 1 ram stick when I paid for 2?

[AverageDude000]

@SUPR3M3 OM3G4 I'm not confident in my abilities to do overclocks/undervolts etc. but I also wouldn't want to because a new card is supposed to work well out of the box isn't it?
I was able to only try one game and currently can't afford more games just to test the gpu on.
Is there any free software I can use to test it?
Also, the youtube issue is more serious I feel like, because it should be running smoothly shouldn't it?
Thank you for your answers.

[Chapunna]

9 minutes ago, AverageDude000 said:

@Chapunna Hi thank you for answering.
Can you please be a bit more specific on how those things will help me?
I have a 1440p monitor, why would I run it at 1080p for example?
Also why use only 1 ram stick when I paid for 2?

Just try it for testing.

[AverageDude000]

5 minutes ago, Chapunna said:

Just try it for testing.

Unfortunately I am physically incapable of trying some of those things (cmos battery, removing ram sticks, checking cables etc.) but If I end up going to a professional I'll ask him to check those things.
Reinstalling windows will be a last resort if I don't find another solution.
Also, isn't Microsoft ending support for Windows 10 in like a year and half?
 

Edited February 20, 2024 by AverageDude000

[Chapunna]

4 minutes ago, AverageDude000 said:

Unfortunately I am physically incapable of trying some of those things (cmos battery, removing ram sticks, checking cables etc.) but If I end up going to a professional I'll ask him to check those things.
Reinstalling windows will be a last resort if I don't find another solution.
Also, isn't Microsoft ending support for Windows 10 in like a year and half?
 

Yes but many will still use windows 10 for a years after end support.

[AverageDude000]

Hey guys I just installed HWiNFO to check if the Adrenalin info is accurate and it says:

"Virtual Machine Warning: Microsoft Hyper-V is active. Some results may not reflect real hardware !"

What does this mean and could it be causing my issues?

[AverageDude000]

Don't know if I'm allowed to bump a thread but might as well try in case any more help comes my way.

[TatamiMatt]

9 hours ago, AverageDude000 said:

Don't know if I'm allowed to bump a thread but might as well try in case any more help comes my way.

You can bump the thread daily so no worries, now, im not too sure about the thing youve mentioned above but i dont think its an issue