PC experiences frequent crashes following the installation of a used RTX 3070 graphics card.
PC experiences frequent crashes following the installation of a used RTX 3070 graphics card.
I bought the RTX 3070 recently, mining for under three months at around 370 euros. It worked well after installation with the latest drivers and DDU. I played for four hours straight without problems, then the PC restarted. After that, it would only run for about a minute before shutting down again. Once I reinstalled the same driver, cleaned everything up, and replaced the GPU, it still only worked for two hours before another issue appeared—a BOSD error. The GPU felt hot to the touch but didn’t overheat or emit any unusual smells. My current specs are a Ryzen 5 2600 (soon upgraded to Ryzen 7 5800X), an ASUS Prime B450M-A, two 8GB RAM sticks, a 2666MHz case fan, and a 750W Alimox power supply. I’ve tried reseating components, reinstalling drivers, disabling DOCP, removing the side panel, and even using an older 1660Ti motherboard to avoid DDU issues. The problem seems to happen after a short gaming session, possibly due to thermal stress or driver instability. It’s possible the GPU is still functional but struggling under load, or the CPU/Motherboard might be reaching limits. I’m wondering if the GPU itself is faulty or if the system is simply too old for optimal performance.
Looking at the thermal paste, it seems possible you might have baked it. Check the guide on fixing video scheduler errors in Windows 10 if you're unsure. Useful resource linked for further help.
I don’t have any extra thermal paste left, and the board seems completely undisassembled.
It seems the temperature suggests a potential problem with the thermal paste or pads. Have you verified that the fans are operating properly? Also, could you confirm the exact model and brand of the GPU? Just a note, that information isn’t directly relevant to the concern. If you can’t return the card, you might want to consider buying one—it’s quite affordable. You may also need thermal pads, but it’s wise to check those first before replacing anything.
Sure, it might become dry or stop working properly... it shouldn’t occur so quickly, though it can. They were also extracting materials from it.
Based on the mining rig they can overheat, which then requires miners to install fans for proper cooling. Many individuals, when discussing crypto mining, have rushed to purchase equipment and set up rigs based on the "gold rush" trend. It wouldn't be surprising if a significant portion of the mined GPUs are in decent condition but far beyond their typical lifespan with TPMs. While cards can handle continuous temperatures between 60 to 70°C or higher, prolonged exposure at those levels—especially around the clock—might cause serious damage. My EVGA RTX 3080 (60C idle auto fan) easily withstands 20 to 40°C under auto load, which is much cooler than my 2070 RTX EVGA FTW3 Super (around 45°C at idle). I can only guess that most 30-series GPUs were used urgently and needed a TPM upgrade.