Seeking another perspective... Part 2: Possibly I'm the issue
Seeking another perspective... Part 2: Possibly I'm the issue
Hi, ten days ago I began discussing the issues with my nearly two-year-old PC. You can find the full post here. I chose to disregard user name filpo who suggested purchasing an 80plus gold thermaltake PSU or a Legion Pro. Instead, I bought an Adata xpg Pylon, which falls within the same price range as the other PSUs. Despite this, I still experience crashes during gaming. I verified that my OS wasn’t corrupted, checked the disks, RAM, and PSU cables, but the problems persisted. After realizing it might be my own mistake, I started repasting the GPU after it overheated, using some fresh thermal pads from an old gaming laptop. I’m unsure if the new pads helped or worsened things. The situation is chaotic after all the work. I’m considering buying more Noctua NT-H2 and applying them carefully to see if it improves performance. Is this mess contributing to the crashes? Could proper application of these pads resolve the issue? Thank you for reading me, and please support me in fixing these mistakes.
It's tough to say if your thermal pads are the problem; you'll need the correct thickness for proper contact with the cooler. For spreading the paste on GPUs, I suggest applying it thoroughly to achieve even coverage. Look up the appropriate pad size and placement to ensure it's not the pads themselves causing the issue.
I’d wipe down the thermal paste just in case it isn’t completely non-conductive. Use isopropyl alcohol or acetone to remove it from the decoupling capacitors and resistors. Apply the paste directly onto the actual die (the central square). Spread a small amount using a credit card or similar tool to cover the entire area. You don’t need premium thermal paste—basic options like Arctic MX-4 (or comparable) will suffice. The variation between pastes is usually 0.1–0.4°C, which isn’t noticeable. The graphics card should self-regulate fan speeds to handle any temperature shifts. RAM pads should match their original dimensions and thickness, and they should be thick enough to press against the heatsink slightly when installed. RAM modules should function normally if they stay below about 80°C. If your system shows high temps, you might check with tools like HWInfo64 or similar software while running a game to monitor temperatures and identify any glitches. The Polaris (RX 4xx) from AMD now supports higher operating temperatures, so chips like the RX6600 can run at 90°C without problems. Their BIOS includes fan curves that favor quieter graphics, which might explain why you faced issues. You could adjust the fan curve in the AMD Control Center or lower the memory clock if temperature-related errors seem linked to memory performance.
The replacement thermal pads, the blue ones, match the height of the ones I previously installed. When securing the GPU, I notice they're secured by the GPU itself, though I'm unsure if this is due to the poor paste application or something else affecting my PC.
I purchased the noctua after watching several videos highlighting its superior performance compared to other brands. However, after nearly two years, I find it quite unusual. This product doesn’t harden like others; instead, it disappears over time and resembles wet sand rather than a dried paste. The high temperatures I experienced—around 85 to 99ºC during games like Cyberpunk 2077 or Mad Max—are similar to laptop temperatures. I wasn’t aware my PC could handle that heat, yet when I opened the GPU for the first time, the original paste was crusty and rigid. Most online guides suggest a safe operating range of 60 to 70ºC for the GPU, which matches the temperature after the initial repaste, but then everything began to fail.