F5F Stay Refreshed Hardware Desktop The friend's computer is acting strangely after the GPU failed, and I'm not sure what might be causing it.

The friend's computer is acting strangely after the GPU failed, and I'm not sure what might be causing it.

The friend's computer is acting strangely after the GPU failed, and I'm not sure what might be causing it.

A
arteralino
Member
137
06-18-2022, 06:27 AM
#1
https://i.imgur.com/w8lbuhL.png We assembled it in 2019 for gaming and video editing. Initially, we planned to use an Intel processor with an integrated graphics card and upgrade later to a better GPU. However, because AMD was rapidly gaining ground at the time, and the Intel CPU would aid his video work, we chose AMD and bought a budget used GPU temporarily. It functioned well until about a month ago when he reported crashes, visual glitches, and YouTube issues. I suggested updating his drivers, which reportedly resolved the problem. A week later he came back with recurring issues. We scheduled a visit to inspect it. Upon arrival, the screen displayed a checkerboard pattern, along with shimmering pixels in games and vertex explosions—clear signs of a video card fault. To be safe, I ran a virus scan, reflashed the GPU's BIOS, updated the motherboard's outdated BIOS, cleaned the card and case with an air blower, and ran the latest memtest86+. Everything seemed normal except for possible GPU failure. We looked for a used GTX 1060 on eBay, which were now cheaper. He contacted me again, now experiencing BSODs during boot and bootlooping, eventually ending up in Automatic Repair. I thought his card was failing completely, blocking Windows from starting, but he needed the PC for work. I suggested a low-cost GTX 720 as a temporary fix. I purchased it on eBay for around $17. After installation, the card worked fine (though performance was modest). I removed the GTX770 to install the GT720, noticed the PCIe slot clip had broken, and reinstalled it. I also found the missing USB-C cable and connected it, along with a working SATA power cable. I plugged in a Blu-Ray burner he’d just bought. I connected the PSU to the case but didn’t find his motherboard box or drive screws, leaving those unresolved. I re-secured the case, checked the BIOS, and ran another stress test with Linux Mint LiveUSB. Booting into it was difficult; it would lock up randomly or enter safe mode, crashing apps. Even after multiple attempts, I couldn’t boot successfully. The system would freeze or behave erratically, sometimes showing error codes like "irql not less or equal" or "page fault in nonpaged area," and other unfamiliar messages. Searching online didn’t yield clear answers. My motherboard’s BIOS appeared correct, so I suspected a possible Windows installation corruption from repeated reboots due to the failing card. I tried another memtest86+ with a Linux Mint USB, which passed. I booted into the LiveUSB for stress testing, but it still struggled to start. Occasionally it reached the desktop, then froze or entered safe mode. I wasn’t confident in using any advanced commands, as my Linux experience was limited. The system would sometimes work when all covers were off, but often required full reassembly. My friend recommended trying the older GTX 770 again, which worked fine in Windows. We considered reassembling everything with the GTX770 until the 1060 arrived, but even then, BSODs occurred during boot with random errors. Now, we’re at a crossroads—do we try to fix it ourselves or wait for replacement parts? No spare components are available, and the warranty on the motherboard expires soon. This situation is frustrating, especially since we were trying to help a close friend. We plan to meet again once the new GPU arrives, but until then, I’m unsure what’s wrong or how to proceed without major hardware replacements.**
A
arteralino
06-18-2022, 06:27 AM #1

https://i.imgur.com/w8lbuhL.png We assembled it in 2019 for gaming and video editing. Initially, we planned to use an Intel processor with an integrated graphics card and upgrade later to a better GPU. However, because AMD was rapidly gaining ground at the time, and the Intel CPU would aid his video work, we chose AMD and bought a budget used GPU temporarily. It functioned well until about a month ago when he reported crashes, visual glitches, and YouTube issues. I suggested updating his drivers, which reportedly resolved the problem. A week later he came back with recurring issues. We scheduled a visit to inspect it. Upon arrival, the screen displayed a checkerboard pattern, along with shimmering pixels in games and vertex explosions—clear signs of a video card fault. To be safe, I ran a virus scan, reflashed the GPU's BIOS, updated the motherboard's outdated BIOS, cleaned the card and case with an air blower, and ran the latest memtest86+. Everything seemed normal except for possible GPU failure. We looked for a used GTX 1060 on eBay, which were now cheaper. He contacted me again, now experiencing BSODs during boot and bootlooping, eventually ending up in Automatic Repair. I thought his card was failing completely, blocking Windows from starting, but he needed the PC for work. I suggested a low-cost GTX 720 as a temporary fix. I purchased it on eBay for around $17. After installation, the card worked fine (though performance was modest). I removed the GTX770 to install the GT720, noticed the PCIe slot clip had broken, and reinstalled it. I also found the missing USB-C cable and connected it, along with a working SATA power cable. I plugged in a Blu-Ray burner he’d just bought. I connected the PSU to the case but didn’t find his motherboard box or drive screws, leaving those unresolved. I re-secured the case, checked the BIOS, and ran another stress test with Linux Mint LiveUSB. Booting into it was difficult; it would lock up randomly or enter safe mode, crashing apps. Even after multiple attempts, I couldn’t boot successfully. The system would freeze or behave erratically, sometimes showing error codes like "irql not less or equal" or "page fault in nonpaged area," and other unfamiliar messages. Searching online didn’t yield clear answers. My motherboard’s BIOS appeared correct, so I suspected a possible Windows installation corruption from repeated reboots due to the failing card. I tried another memtest86+ with a Linux Mint USB, which passed. I booted into the LiveUSB for stress testing, but it still struggled to start. Occasionally it reached the desktop, then froze or entered safe mode. I wasn’t confident in using any advanced commands, as my Linux experience was limited. The system would sometimes work when all covers were off, but often required full reassembly. My friend recommended trying the older GTX 770 again, which worked fine in Windows. We considered reassembling everything with the GTX770 until the 1060 arrived, but even then, BSODs occurred during boot with random errors. Now, we’re at a crossroads—do we try to fix it ourselves or wait for replacement parts? No spare components are available, and the warranty on the motherboard expires soon. This situation is frustrating, especially since we were trying to help a close friend. We plan to meet again once the new GPU arrives, but until then, I’m unsure what’s wrong or how to proceed without major hardware replacements.**

B
badgergirl0315
Member
163
06-18-2022, 06:27 AM
#2
B
badgergirl0315
06-18-2022, 06:27 AM #2

X
xISplex
Member
221
06-18-2022, 06:27 AM
#3
It was fastened using two screws. There aren’t any other x16 ports, but I used an x8 slot with the GT720 without any change. He also has the SATA cables and screws needed for that optical drive now, so next time I can simply load it from a DVD-RW. I can use the "toram" feature to transfer the whole OS into RAM and avoid any problems with the drive.
X
xISplex
06-18-2022, 06:27 AM #3

It was fastened using two screws. There aren’t any other x16 ports, but I used an x8 slot with the GT720 without any change. He also has the SATA cables and screws needed for that optical drive now, so next time I can simply load it from a DVD-RW. I can use the "toram" feature to transfer the whole OS into RAM and avoid any problems with the drive.

X
xAdriLCT
Senior Member
702
06-18-2022, 06:27 AM
#4
Well, things are getting even more puzzling. I set up the PC at home and started by unplugging the SATA drive, then checked the BIOS settings. Everything looked fine, so I ran the newest beta of Memtest86+. Errors appeared. I swapped the RAM without success—constant crashes followed. I rechecked the BIOS and saw an "auto" option for CPU overclocking that was probably turned on by default. I mentioned updating the motherboard BIOS, knowing it could break any overclocks even though we hadn’t manually adjusted them. I tried turning off both the auto setting and XMP, which stopped the issues. After disabling them, I let the system run for over 15 hours with multiple Memtest86+ runs—no errors. To confirm, I also ran Memtest86 free, which showed no problems. For my relief, I updated to the latest Memtest86 beta, but it kept throwing nonstop errors. I rebooted just to verify, and still got the same issue. After a power cycle, the BIOS reset back to defaults. I re-enabled XMP and tried disabling the auto overclock setting, but it warned me about potential instability before I acted—so I chose not to. Now Memtest86 beta is passing, yet the system keeps crashing unexpectedly after nearly 24 hours of RAM testing. This sudden shift from stable to erratic behavior has left me even more baffled than before.
X
xAdriLCT
06-18-2022, 06:27 AM #4

Well, things are getting even more puzzling. I set up the PC at home and started by unplugging the SATA drive, then checked the BIOS settings. Everything looked fine, so I ran the newest beta of Memtest86+. Errors appeared. I swapped the RAM without success—constant crashes followed. I rechecked the BIOS and saw an "auto" option for CPU overclocking that was probably turned on by default. I mentioned updating the motherboard BIOS, knowing it could break any overclocks even though we hadn’t manually adjusted them. I tried turning off both the auto setting and XMP, which stopped the issues. After disabling them, I let the system run for over 15 hours with multiple Memtest86+ runs—no errors. To confirm, I also ran Memtest86 free, which showed no problems. For my relief, I updated to the latest Memtest86 beta, but it kept throwing nonstop errors. I rebooted just to verify, and still got the same issue. After a power cycle, the BIOS reset back to defaults. I re-enabled XMP and tried disabling the auto overclock setting, but it warned me about potential instability before I acted—so I chose not to. Now Memtest86 beta is passing, yet the system keeps crashing unexpectedly after nearly 24 hours of RAM testing. This sudden shift from stable to erratic behavior has left me even more baffled than before.