Game pauses unexpectedly on Discord while PC experiences crashes.
Game pauses unexpectedly on Discord while PC experiences crashes.
Afternoon, I've experienced occasional shutdowns over the past two months and am struggling to pinpoint the cause or figure out how to fix it. My suspicion points toward the motherboard, RAM, or processor since I upgraded them in April, but the system remained stable from April until late August. Early Signs - Discord Desktop issues Around September I noticed Discord would crash during gameplay, while streaming and voice chat were active. Opening OBS afterward would cause it to freeze and take several minutes before it could be launched again. The problem has become more consistent: Discord often crashes when using OBS alongside a game in under five minutes. As a workaround, I switched to web access for Discord, but that removed voice chat functionality, which is inconvenient. I haven’t found any relevant entries in Discord’s logs, cleared the cache at %appdata%/roaming and %appdata%/local, and performed full reinstalls—still no improvement. Second Symptom - Applications crashing I've encountered game crashes at random moments during interaction or loading. It affected OBS, Genshin Impact, Baldur's Gate 3, Elder Scrolls Online, Cyberpunk, and Starfield (except Starfield), which were usually stable before. The interfaces in games would freeze if I rushed to restart instead of allowing time for reloads. Errors appeared as NVEC warnings, page file issues, and vague messages, but nothing specific. This occurs less frequently than Discord problems but happens consistently when it does, often requiring patience until things stabilize—especially during streaming. Third Symptom - PC crashes This happens after opening the system for a while, especially when switching between full-screen games and OBS. It occurred with Genshin, Baldur's Gate 3, Elder Scrolls Online, and Starfield, though generally stable otherwise. Errors in UI elements and other messages appeared if I didn’t wait before relaunching. My current setup: Motherboard – MSI MAG Z790 TOMAHAWK WIFI CPU – Intel Core i7-13700K GPU – ZOTAC GAMING GeForce RTX 3080 Trinity OC White Edition Memory – G.Skill RipJaws S5 Series (Intel XMP) 32GB (2 x 16GB) 288-Pin SDRAM DDR5 6400 CL32-39-39-102 1.40V Dual Channel SSD1 (OS and main games) – SAMSUNG 980 PRO SSD 2TB PCIe NVMe Gen 4 Gaming M.2 SSD2 (storage and other games) – ADATA XPG SX8200 Pro 1TB 3D NAND NVMe Gen3x4 PCIe M.2 2280 Power Supply – CORSAIR - RMx Series 850W ATX12V 2.4/EPS12V 2.92 80 Plus Gold Modular Power Supply OS – Windows 10 Tests Run: Memtest86+ (no issues), DirectX Diagnostic Tool (no problems), Furmark (preset 1080, 1440, 2160, GPU stress) – all passed; CPU-Z and MSI Bench tests also completed without errors.) I have tried uninstalling Discord, clearing the cache, reinstalling in safe mode, running driver uninstallers, flashing the board, updating drivers, updating Windows, and monitoring crash logs. Mostly hit 100% usage on CPU, memory, or disk, with temperatures normal. I’m currently planning a full wipe and a fresh Windows 11 install, but wanted to see if any additional tests or overlooked signs exist. Appreciate your assistance.
Visit C:\Windows\Minidump to see if any minidump files exist. If present, return to the Windows directory and transfer the entire Minidump folder to the Downloads folder (use your desktop if needed). Compress the copied folder and include it in a post. Please adhere strictly to instructions since Windows doesn't allow file manipulation there.
In light of the same problems you're facing, I've tried similar solutions as you did. One suggestion I found helpful (a Reddit user recommended) is to install an older version of the Nvidia drivers—starting with 9/12/2023 version 537.34. If that doesn't work, uninstall it and try a previous, even older release, repeating the process until you find a stable one.
The logs show multiple crashes: ntoskrl.exe stopped 6 times due to index mismatch, PFN_LIST_COPRRUPT once, PAGE_FAULT_IN_NONPAGED_AREA once, blank errors twice, HoyokProtect.sys failed once with a kernel trap, UNEXPECTED_KERNAL_MODE_TRAP, NTFS.SYS crashed once with a page fault, and win32kbase.sys ended with a system service exception.
I might attempt it, though I'm concerned about stability if I use an older driver alongside newer games.
RAM appears in the dump files. If you've performed any overclocking, revert it back. This covers the RAM frequency as well—5600MT/s is the maximum officially supported rate for that processor. If the system remained stable for over a month after installation, the issue might not be there, though it's simple to verify. For checking RAM, I avoid memory testers because they often miss faulty units. Instead, I run the machine normally with one stick at a time and observe which one causes crashes.
Using an odd count of sticks isn't usually advised. If something is faulty, rely on the warranty. RAM comes with a lifetime guarantee. They typically require the whole set to be returned, so for a fresh purchase it's best to keep only the new kit. You might try running three sticks, but stability can be an issue. The full return policy exists because they often change suppliers between production runs, making batches distinct.