Problem Occurs with Updated Stable GPU
Problem Occurs with Updated Stable GPU
Hello!
Recently I encountered issues with my GPU being overclocked in games.
My setup includes:
EVGA GTX 1070 FTW
i5 6600K OC 4,5Ghz
16Gb RAM
MSI Z170 Gaming M5
Driver - 452.06
Last month I had to reinstall my operating system, Windows 10 x64 Home edition. Before that – everything ran smoothly and the GPU overclock was set via MSI Afterburner with specifications of +150Mhz core, 112% power, and a temperature limit of 88°C.
With this configuration, I didn’t need to increase the voltage. In games like RDR2 and Witcher 3, using the same overclock settings worked fine, and I noticed improved FPS.
After reinstalling the OS, I stopped playing those games and didn’t perform any further overclocking. Recently, I wanted to replay RDR2 and recreated the previous overclock preset in MSI Afterburner. However, whenever I launch a game with an OC preset (even at 20-50Mhz or higher cores), after loading and starting the gameplay, it crashes with errors related to the GPU.
Specific problems:
- RDR2 – after loading, lots of artifacts appear and the game crashes in 1-2 seconds with ERR_GFX_STATE error.
- World of Warcraft – crashes immediately after loading with "Error 3D-accelerator".
This issue occurs with both the 452.06 and 451.67 drivers, and it happens regardless of whether I enable Windows 10 features or not.
Anyone else experiencing this? Could you help me understand what’s going on?
Thank you in advance!
The artifacts, mistakes, and crashing sounds resemble a failing GPU.
But without OC at all – the GPU performs perfectly, no issues, no artifacts, no freezes, smooth FPS. However, adding a minimum 50Mhz boost on the core causes it to crash. Before reinstalling the OS, I’ve never seen anything like this even close...
My 1070 runs smoothly at 1976Mhz without extra manual tuning, especially when cold (from Russia 😄). It actually reaches higher MHz.
GPU didn’t overheat much—between 70-72°C, maybe briefly reached 76°C in Furmark before stabilizing with a good fan setup, but that was a long time ago.
How can I tell if it’s failing?
I haven’t tested earlier drivers yet; I’ll try today after work, but if the GPU is dying, I don’t want to scare you.
I'm going to perform DDU and apply the prior driver to check for updates. Nvidia support mentioned that this problem might stem from new drivers or could involve HAGS.
Well,
With HAGS on, after DDU and OC preset +100-120Mhz (power at 112%, temp around 88°C) everything functions well without artifacts or stuttering, and FPS improves. +130 mhz causes artifacts to appear, while above 140mhz there are many issues. Even when adjusting voltage via afterburner (+2-5%), the voltage remains unchanged—maximum 1.062v, just like with default EVGA OC. The preset is stable at +100mhz, GPU operates smoothly, and it stays at 2088 mhz (base EVGA OC around 1978Mhz). Temperature stays below 70°C (66-68°C stable). I could try up to +120Mhz, but I don’t want to reduce GPU usage too much.
I’m not sure why, but it seems the GPU didn’t exceed 2050mhz before OS reinstall—remembering that it would drop about 50mhz once it got hot (around 70-72°C).