Artifacting and crashes with the 6750 XT – I’ve done everything possible
Artifacting and crashes with the 6750 XT – I’ve done everything possible
I'm struggling with my friends' PC specs: Ryzen 7 7700X, MSI RX 6750 XT, 12GB OC, 32GB DDR5 RAM. It was assembled in December 2023. After a few months away fighting fires, he returned with issues. Windows showed artifacts on the GPU, so I sent it back for replacement. When it came back, problems persisted. Recently, after replacing parts and updating drivers, it still crashes during intense usage—especially in dark or darker scenes, even when playing games for over five minutes. I've tried various fixes: new PSU, motherboard, RAM, driver updates, BIOS updates, and even changed fan settings. The temperature never spikes enough to trigger a crash. I've also considered the possibility of faulty power supply in his home, which might be affecting the system. The situation feels frustrating and unresolved.
I understand your frustration. Here are some steps to consider: 1. Test the 6750 XT in another known working system to see if the issue persists. 2. Check if the Mobo has a second PCIe port and try using it there. 3. Confirm the RAM setup is 32GB DDR4 2x16. Try a different slot and enable XMP/EXPO settings. 4. Verify if any CPU or GPU modifications are in place. Run everything at default settings. 5. Determine whether you're using Windows 11 or 10 and try the alternative version. 6. When the system crashes, does it force a hard crash requiring you to press the power button for four seconds, or does the driver simply crash and the screen goes dark? The PC stays on but no display appears. When you restart into Windows, look in Event Viewer under Windows Logs > System for any relevant entries. As an Nvidia specialist, a driver crash typically shows up there with details similar to the image provided. Generally, the final part is unreliable, but the screen may stay blank while audio continues playing from speakers as the machine doesn’t completely freeze.
Do you have a video of the items? Usually you can tell if it's memory or core related. You might want to reduce the memory clock and see if that improves things; if it does, the problem remains. Don't rely on xfx to fix it if they can't locate the issue—they'll send you the card back. The more details you share, the easier the RMA process becomes.
Here’s a revised version of your notes:
1) Used the second card before initiating RMA; it caused my PC to crash, but this unit had a fresh serial number so it didn’t. Now tried it and it crashed immediately on my new PC. No visible artifacting, but it did crash with an AMD driver error warning similar to what appears on his machine.
2) Tested on another PCIe slot on his PC – same crashes happened.
3) Installed brand new RAM (2x16), tried both default and EXPO 1 mode – still experienced crashes.
4) No overclocking performed; everything started at stock settings.
5) Began with Windows 10, upgraded to 11. The latest attempt went straight to 11 but still triggered crashes on both systems.
6) No overclocking at all; everything remains at original specifications.
7) Started from Windows 10 and moved to Windows 11. This most recent try ended up crashing again, sometimes randomly or after a reboot. I’ve attached a screenshot of the Event Viewer log after a crash, as well as a benchmark report showing artifacting in the top left corner.
I submitted the card again and they confirmed a brand new one was sent. I tried it at home and successfully played for an hour and a half without any problems, even running a full game of helldivers. Later, when I brought it to his place, I encountered issues right away. Here are the screenshots of the successful long tests on his PC before it arrived.
He had a problem because his GPU caused his PC to crash. I used an old 970 chip in his system and it functioned smoothly. I plan to return it. If the issue returns at his home, I’ll blame the power supply there. Otherwise, I’ve already been scammed on two RMAs and three MSI 6750XTs.