No, your GPU is not defective. It appears to be functioning properly.
No, your GPU is not defective. It appears to be functioning properly.
Hi there, a few months back I purchased the components for this setup and had a tech expert assemble the PC. Since then, I've encountered issues with the graphics card. While playing a game, I experienced frequent random crashes and stuttering. The temperatures climbed to 95°C, so I crafted a custom fan curve in Adrenalin—after that, the heat stayed manageable, reaching a maximum of about 80°C. I also increased the GPU power limit to its highest setting (15%) based on some advice. After restarting, the problems persisted, and one crash forced me to shut it down via the power button. Upon rebooting, the card was disabled in Device Manager. I tried reinstalling drivers in safe mode using AMD Cleanup Utility, hoping it would improve things. That night, while closing Windows to turn off the PC, another crash happened and the settings reset—consistent with previous crashes. I didn’t play games then; it crashed outright when I pressed Shutdown. This time I switched to DDU for the graphics drivers and updated the chipset drivers since they were outdated. I also planned a clean game reinstall but ran into more problems beforehand. While watching a YouTube video, it stuttered (audio and video). With Adrenalin active, during stutters GPU utilization, clock, and voltage dropped to zero. I also observed the GPU memory clock stuck at zero for several minutes afterward. I’m using Adrenalin to monitor these metrics—could they be inaccurate? I’m currently streaming on Firefox.
Additional details: OS is Windows 11 Pro 23H2. I updated the BIOS shortly after purchasing the PC and enabled XMP without further tweaks. I added an extra intake fan at the front (now two intakes front, two exhausts back and up).
Where should I begin troubleshooting? Might one component be faulty? Should I run DDU again? Perhaps a fresh Windows install would help?
Thanks in advance for your guidance—I’m trying to understand this better.
Best,
Edited February 20, 2024 – AverageDude000
The temperature of 95°C is quite high. Here are some quick questions: Was the card brand new or second-hand? How do the temperatures look on your other parts, such as the CPU? What is the average room temperature in your space? Did you open the card to install the thermal paste? Which case does your card and PC come in, and what fans are installed? Were there any other GPUs running on your current Windows setup before adding this one? Could you also provide a photo of your PC now? Show me where it sits on your desk and how it looks when you use it? With these details, we can better diagnose the issue.
1. All components are brand new.
2. CPU stays at 40 °C when idle, reaching a maximum of about 55 °C while gaming.
3. Room temperature averages between 22 and 25 °C.
4. I purchased the parts right away and brought them to a technician because I couldn't assemble it myself. No modifications were made.
5. The case model is a "be quiet! Pure Base 500DX ATX Mid Tower." I only added an extra intake fan at the front; the current setup has two intakes (front) and two exhausts (back and up).
6. The existing PC parts are unchanged; I haven’t replaced any components, and the machine is only a month old.
7. I can't post a photo yet, but the case is on the floor with enough clearance. The hottest spot reached 95 °C during gaming before installing the custom fan curve. I suspect Adrenalin was struggling, and the Zero RPM setting didn’t kick in properly—fans only started around 58–60 °C.
Update the bios information. Remove the CMOS battery, then reinsert it. Verify the GPU power cable connection is correct. Reinstall a brand new Windows 10. Disable secure boot settings and install an outdated GPU driver (4–5 months old). Make sure to download all drivers from the motherboard manufacturer's website.
You're asking for advice on troubleshooting your GPU issues. Have you considered adjusting the voltage and lowering the clock speed? This could help identify if the problem lies with the chip or its performance. If the issue persists, a clean Windows install with updated drivers might be necessary. If that doesn't work, it could indicate a faulty GPU. Also, would you like to know if the crashes occur only with certain games? For instance, your PC struggles with one specific title due to a known 14900K issue, while developers are addressing it.
Certainly! Here’s a clearer version of your questions:
@Chapunna Thank you for your response. Could you explain more about how these changes will benefit me? For instance, with a 1440p monitor, why would I choose 1080p instead? Also, why should I use only one RAM stick when I bought two?
I’m unsure about handling overclocks or undervolts myself, but I also don’t want to risk it since a new card is meant to perform well right out of the box, isn’t it? I managed to play just one game and can’t spare more just to check the GPU. Are there any free tools you recommend for testing? Also, the YouTube problem seems more significant—I really hope it runs smoothly. Thanks for your help.
I can't perform those tasks myself, but if I go to a professional, I'll have them inspect those things. Reinstalling the windows will only be considered if nothing else works. Also, Microsoft is stopping support for Windows 10 around next year and a half. Edited February 20, 2024 by AverageDude000