Driver issues causing repeated crashes on NVIDIA hardware
Driver issues causing repeated crashes on NVIDIA hardware
Accidental message sent because of the outage. Updated on November 29, 2021 by Heavybell
It looks like you’ve tried many things without a clear path forward. Here’s a concise summary: Restore your BIOS to default settings—don’t alter stock configurations unless essential. Reinstall Windows from the beginning. Install GPU drivers and check chipset compatibility. Test components one by one, especially if multiple drives or peripherals are connected. Verify RAM usage and consider swapping it out temporarily to see if the problem persists. Focus on controllable factors and clean up your setup before proceeding.
I've reinitialized the BIOS to help with the 3080 issue. I've adjusted some configurations but haven't reset everything fully yet. Windows reinstall remains on my list of options this week. I prefer avoiding a full reinstall; instead, I'm focusing on restoring things to my preferred settings—like deleting zip folders and unwanted UWP applications, updating software, and ensuring GPU and chipset drivers are current. I haven't tested any RAM sticks yet, but I've experimented with different configurations. Recently, I got new NVMe drives and plan to move games onto one, remove the Gigabyte AIC card, and disable AMD RAID. I'll keep RAM at normal speed (around 2166 MHz) and see if that helps. I'll also check the single monitor setup before considering a reinstall.
Take care of yourself by removing the existing NVMe drives and installing Windows from scratch on the new ones to confirm the problem persists. You've already invested a lot in other projects, so reinstalling and setting it up would have been more expensive. In reality, this is essentially the last resort—everything else has been swapped out.
Alright, quick update on this. At a suggestion from another board, I am currently running the RAM without XMP enabled (so 2166 MHz I think). Because of my own suspicion I removed the Gigabyte AIC card, replacing it with a pair of Seagate FireCuda 530s. I've disabled PCIe RAID and reinstalled Windows 10 from scratch using newly created media. Finally, I installed the chipset drivers from AMD and the 496.76 DCH GPU drivers. This was a few days ago and so far, no problems (knock wood). I'm going to keep an eye on it for another week or so before declaring it fixed, and if things go well I'll turn RAM XMP back on. There's also a bunch of software I've not installed yet that could potentially cause issues, e.g. iCUE for my keyboard and the G.Skill RGB controller for my RAM (I like to turn the lights off. At the time I bought it there was no similar speced RAM on the CPU's tested list that I could feasibly acquire that didn't have RGB on it). Edit: Hm. My main monitor just did the 'go blank for a second or two' thing that I thought had disappeared in the reinstall, that's worrying...
It has been a while since my GPU driver encountered an issue, so I'm revisiting this for anyone who might search for it later. Eventually, I restored the RAM speed to 3200 MHz with XMP enabled. I’m considering keeping it at that setting instead of reverting to 3600 MHz, though I’m not sure. I’ve reactivated Precision Boost Overdrive. The Gigabyte AIC remains outside the build, and I have a brand-new Windows 10 installation with PCIe RAID disabled in the BIOS (switching from RAID to AHCI without a reinstall seemed impossible; I tried various workarounds). I’m ready to accept that the fix came from the reinstall, but I’d rather not risk it now. By the way, I encountered another issue with frame rate drops in some games, which happened after installing Windows Sandbox. From what I found online, it might have been related to virtual network adapters. I only use Sandbox occasionally, so removing it isn’t too bothersome. I haven’t installed iCUE or the G.Skill RGB controller yet. If I do and the problem returns, I’ll make sure to reference this thread. Thanks for all the guidance.