The most haunted computer ever recorded.
The most haunted computer ever recorded.
I believe I own the most troubled computer I've ever had. I assembled a machine in 2022, an all-AMD configuration, expecting a stable performance for 1440p. I had no idea it would become a never-ending struggle of anxiety and fixes to keep it running. One day while playing Helldivers 2, the GPU abruptly failed. Usually I’d restart it easily, but today it took over five minutes to boot, a situation I definitely don’t want again because it’s expensive.
About this machine, I’ve faced frequent BSODs almost every month, and sometimes the boot device wouldn’t even appear. When it did work, it was a miracle. Recently I saw my settings changed unexpectedly—despite having updated drivers—I’m confused about what happened. I panicked and tried a Q-Flash, which somehow got me back online.
Now I’m unsure what’s coming next. My worry is that this could become a costly issue, and all I can do is hope it stays functional and try to patch it up.
I own a Ryzen 7 5700X, Gigabyte RX 6750XT OC, 16x2 Corsair Vengeance 3600 MHz CL18, a Kingston NV2 TB (boot drive) plus a 2TB Team Group MP33, and a Corsair RM850x PSU. When it comes to fixing issues, I usually look up online whenever I encounter BSODs with specific error codes. Since the problem persists despite updates, I often focus on GPU drivers, which I can’t adjust directly. For the boot drive failure, I checked its health but found nothing. Regarding the recent issue, a friend suggested it might be related to the GPU or drivers, and I confirmed I have the latest drivers. I tried using Event Viewer but struggled to interpret the results. I also recalled a post mentioning BIOS settings and a reset, which led me to try that solution.
The changes were made to the hardware configuration, specifically BIOS settings. This reset behavior after Q-Flash is typical. Additionally, Heldivers experiences problems with AMD GPUs when they exceed GPU usage limits, leading to crashes—avoid pushing the GPU beyond full capacity.
Always mention that after a BIOS flash or CMOS reset, we'll examine the dump files from the BSODs. Look in C:\Windows\Minidump for any minidump files. If found, return to the Windows folder and move the Minidump folder to the Downloads folder (use your desktop if needed). Compress the moved folder and send it in a message. Follow these steps carefully—Windows doesn't allow changes there.
Warning: this will be a lengthy and confusing entry about the same black screen we saw today (actually yesterday at 12:50 AM here), along with my RAM troubleshooting experience. TL;DR is at the bottom. For clarity, D1-4 refers to the RAM slots (chosen for simplicity), and I’ll highlight D1-4 in bold while keeping the text readable. The *sticks* are emphasized in italics. After another crash and possibly three hours of intense troubleshooting, I can confidently say my GPU was reaching its limits during a Helldivers 2 session. Another concern was whether my RAM modules might be damaged—because when I checked, both sticks failed when only D2 and D4 were left, and even when I swapped them back, the issue persisted. What worried me was that leaving Stick 2 in D1 caused a boot failure, and when I swapped it to D1 with Stick 1, it would turn on and off repeatedly. Knowing the AM5 DDR5 timing requirements, I suspected a timing mismatch rather than a simple corruption. I tried putting both sticks back in their original slots after booting, but the screen still went black. Thinking it was Stick 2, I removed it from D4 and tried again; it worked. After that, I swapped Stick 1 into D2 & D4 and everything functioned normally. Earlier, when I put Stick 2 in D1, Stick 1 worked fine, which confused me. It seemed both sticks could operate even when both were in D1 & D3. I finally ran a test on D2 & D4 to see if I needed to return the motherboard (still under warranty), and everything worked again—frequency reset to default. That was a simple fix. TL;DR: I spent three hours trying to get my PC running with random working RAM sticks, and now I need advice on limiting the GPU usage to 80% without modifying the BIOS (OC won’t help). I also want to maintain good performance for streaming and recording, with a monitor resolution of 1440p at 75Hz.
Here are some adjustments that resolved the diving problems. I limited the game to 60 or 75 frames per second, as you prefer. Please adjust the GPU clock speed if needed—feel free to walk through it with me. Also, let me know the GPU voltage. Raising the power limit can help too, since it lets you push the GPU harder without hitting maximum usage, which seems to be the main cause of crashes.
I believe the game is limited to 75 frames per second. For the GPU voltage, I checked during my time on the ship and it was at its maximum of 1200mV. The GPU clock speed was 2688 MHz with 92% utilization. I want to understand how to adjust the GPU settings manually, since auto-tuning in Adrenaline Software isn't functioning. Edit: I managed to do it myself—just need guidance on the correct values.
I adjusted the power limit to +15% and set the GPU clock speed to 90%. I also tweaked several in-game settings to reduce usage to about 50-60%, even during missions. The biggest improvement came from switching from Native sampling to Ultra Quality, which significantly cut usage with little effect. I spent an hour without any crashes, so I believe this should work well for Helldivers 2 now. If anything is missing or I can add more, I’ll update the post.