Recent updates have caused texture problems in several games, along with BSOD errors and other malfunctions.
Recent updates have caused texture problems in several games, along with BSOD errors and other malfunctions.
Following a tough start from zero, I've faced several problems, mostly with visuals. Sometimes textures get swapped with error images. First spotted in PUBG (screenshot attached, waist area), then also in Forza Horizon 5. There was a crash with SYSTEM_SERVICE_EXCEPTION for nvlddmkm.sys. After the incident and another after reboot, the screen displayed "Previous overclock settings failed." I didn't adjust anything else except updating the XMP profile in BIOS. Once the message appeared, the XMP settings reset. When the PC powered off, I couldn't activate it with the mouse or keyboard; only pressing the button on the case brought it to life, but nothing showed up. A BIOS update fixed the display issue, though waking from input remained problematic. Mouse and keyboard work in Device Manager. Some attempts: BIOS update, driver reinstall, graphics card placement during build, and driver updates. During setup I struggled to properly install the graphics card. In PUBG temperatures stayed around 60°C (monitored via MSI Afterburner). Hardware specs: MSI MPG Z690 CARBON WIFI, ATX motherboard, BIOS v7D30v13, Intel i7-12700K, etc. I've used Corsair RM850x, Samsung NVMe drives, and Windows 11 Home.
Have you checked the BIOS for a fail-safe configuration? Because you're experiencing BSODs beyond just visual glitches, it’s likely a RAM issue. Running memtest86 might help identify problems. https://www.memtest86.com/
Thank you for your input! The test lasted longer than anticipated (about 4 hours), yet no problems were detected. Attached is the report.
Use DDU for a full driver removal and reinstallation of the specified file. For testing, leave XMP disabled to check if it resolves the BSOD problems. If not, remove the PCIe graphics card and try Intel's igpu; see if it can be put to sleep and wake up. Let me know your results.
The DDU didn't assist effectively. Identical issues persist across games, with a more severe graphics glitch now detected. A screenshot is included. No BSODs beyond the first occurrence were reported. Every restart after using DDU or adjusting settings showed failed OC configurations (screen active) without changes. I suspect removing the graphics card might be necessary. Would it be worthwhile to try it in the PCIe 8 slot? Should I also attempt using my old GTX970 on this system?
Consider testing your GTX970 on this machine. Usually, if a PCIe lane is an issue on one CPU, the others might share the same problem, even though it wouldn't hurt to try another slot. More importantly, if you have the original system, you could run the 3070ti there to determine if the card or the CPU is the cause.
I've adjusted the setup by moving the 3070TI to PCIe 8, but the issues persist over time. I placed it in an older system and it worked perfectly—no problems. I attempted to run demanding games as much as possible, yet problems still arise. It seems the bottleneck might be more related to the CPU and storage rather than the GPU. When playing Forza, I encountered streaming bandwidth issues before, but those never appeared on my new machine. It appears the GPU isn't the sole cause. I've moved the GTX970 to the new system, and it also runs smoothly. Still, I'm struggling to resolve this.
Insert the 3070TI into the new PC. It functions initially but after a couple of hours it returns to graphical issues. These problems don’t hinder general performance, only the visuals. Temperatures stay consistently between 60 and 62 degrees.
Do you have any GPU software like EVGA Precision X1 or MSI afterburner? It seems there are GPU memory errors. If you're using software, ensure it isn't overclocking the memory—this could be the cause. Memory should be at around 9500 stock. If none is running or you had software installed, close it and download Precision X1 from EVGA's site, setting the memory clock offset to -1000. It should then run at about 9502 stock. If not, there may be a problem. Also, note that EVGA might update firmware during startup, which will notify you if an update is happening. Once the clock is set to -1000, it should drop to around 8502, potentially lowering further under load.
I noticed the issues began after using MSI Afterburner for temperature monitoring. It was initially set to 9502. After installing Precision X1 and updating the firmware, I observed fewer errors. However, occasional crashes and new error messages about "Out of video memory" appeared. The problem persisted even with clock offset adjustments. Based on these findings, it seems the issue might be related to the card model.