The NVIDIA GeForce GTX 1060 with 3GB RAM is experiencing crashes and instability during gameplay.
The NVIDIA GeForce GTX 1060 with 3GB RAM is experiencing crashes and instability during gameplay.
My older pre-built machine has been experiencing crashes over the past few years, especially when under heavy use like gaming and multitasking. The display often goes dark before coming back, but the game freezes until I manually terminate it in Task Manager or it shuts down on its own later. I’ve noticed Windows beeping sounds similar to unplugging a device. System logs show errors such as "nvlddmkm" and messages about the display driver stopping and recovering. I’ve attached screenshots of these entries. I’ve used DDU to reinstall drivers, swapped out the SSD (without data loss) for other reasons, and run a RAM test via Command Prompt, though I’m not sure what kind. The only reliable fix so far is adjusting core and memory clock speeds with MSI Afterburner—about 150mhz each. This isn’t guaranteed and sometimes fails, especially in games like Overwatch 2, where crashes still happen occasionally. What steps can help resolve this or postpone the need for a full rebuild by roughly three months? My specs include an Intel i5-7500 CPU, 16GB DDR4 RAM, GTX 1060, 1TB Samsung SSD, and a PSU rated around 650W (though I’m not certain). The motherboard is an MSI B250M MORTAR (MS-7A69).
You're totally new to overclocking! XMP is a great starting point because it's simple. To learn more, you can explore online tutorials or guides that walk you through the process step by step.
In MSI Afterburner there are typically four sliders: GPU core clock, VRAM clock, power limit, and fan speed. It seems you attempted adjustments on the first two, but I recommend modifying the third instead. This change will affect clock speeds differently, potentially improving stability even if performance drops slightly. Keep in mind that disabling certain options might link them together, so ensure you uncheck any related settings if you want to focus on adjusting just slider three.
This was the intended setting. My understanding was slightly incorrect; the tick box displays a chain icon labeled "ON," which controls power and temperature limits together, not separate options. I missed the temperature restriction earlier. It would help to test the lowest levels and see if stability returns. If that doesn’t work, it’s probably not worth adjusting further. If it does, you can experiment with higher settings to balance performance and stability. The achievable limits vary by GPU model and manufacturer—typically around 40% to 70%. To summarize, reset the core and memory clocks to their standard values, disable power and thermal limits only if necessary (one should change), lower the power limit to its minimum and check stability, then explore higher limits for better performance.
I'll rely on your decision but it might help to verify it first. If not, I'm just making an educated guess.