Facing issues? Reboots are unpredictable and occasional BSODs occur during gameplay.
Facing issues? Reboots are unpredictable and occasional BSODs occur during gameplay.
I recently upgraded my machine with a brand new R9 5900x and RM1200x, plus a used 3090 I acquired from a reliable source (he owned it for about four years). The issue began after I reinstalled Windows, which is something I usually do when changing computer components. Frequent crashes occurred in demanding titles needing over 70% GPU usage, like Vermintide 2 and Ghost of Tsushima. Performance holds up well in games such as Skyrim and Assassin's Creed 3 that require less GPU load (around 30-40%). When it fails, the screen freezes for about three seconds before the system restarts. It doesn’t fully crash; it just restarts intermittently. Over two days, this has happened roughly ten times. The first incident triggered a BSoD for half a second, though I didn’t notice any error messages. My most recent failure also displayed a BSoD after disabling Windows AutoShutDown. Despite running for hours without restarting, the machine randomly shuts down after about 30 minutes of gaming.
My hardware specs:
- Old: R7 3700x | B550 Elite | 32GB DDR4 3600MHz Vengeance | RTX 3060 Ti | 1TB Samsung 970
- New: R9 5900x | B550 Elite | 32GB DDR4 3600MHz Vengeance | RTX 3090 | 1TB Samsung 970 | 2TB SN770 | RM850x
I also own a brand new H115 AIO (December purchase) that I bought because my previous one failed.
Running setup:
1. Switched power supply – initially thought insufficient power, but now it works.
2. Eventlog shows WHEA 46, possibly indicating memory issues.
3. Ran MemTest for four hours; results were clean.
4. Replaced RAM, assuming a possible bump during installation.
5. All drivers are current; GPU, CPU, and Realtek are up to date.
6. Used Prime95 for two hours, showing 100% usage without errors.
7. Benchmarked RTX 3090 on Heaven Benchmark – no problems despite high load.
8. Monitored temperatures: CPU ~82°C, GPU ~78°C, RAM ~46°C during testing.
9. Tried disabling AutoWindowsRestart in advanced settings; only a brief crash occurred with Vermintide 2.
10. Updated BIOS to version F19d before adding the Ryzen 9.
11. Tested X.M.P and disabled it – no difference observed.
12. Checked online forums; many users report similar issues but varied fixes.
I’m trying to resolve this by reinstalling Windows just in case it’s a software issue, but I’ll wait before making further changes. The task manager was used on a second monitor to check for unusual activity during crashes, but no anomalies were found. Thank you for your help!
WHEA_UNCORRECTABLE_ERROR is usually indicative of a hardware failure. I ran into this issue when my PSU was dropping voltage on a rail out of spec randomly. You said you already tried a different PSU though and 800W was more than likely enough for the setup if it was good quality. CPU's are normally robust but it's not off the table. My first goto would be to plug whatever old GPU you were using before back in and run the system for a while play some games. A thought entering my mind is how heavy is the 3090? It's not off the table that the GPU could be sagging in the slot enough to where the temperature increase is causing a bad connection. As silly as it sounds you could try laying your PC on it's side, reseat the GPU, leave it on it's side and see if the problem persists. That'd be an easy fix.
Both PSUs and the GPU are from Corsair, and the setup includes a triple fan, making it quite heavy. I added a metal bracket from the box to help support it. I may need to reinstall my old CPU to check for errors... All Windows logs indicate a memory-related fatal hardware error with details available. Many users report issues with other components besides memory. Thanks for your help—I'll investigate further today.
I aimed to reduce the GPU's power consumption to 80% and observed the trend. If the pattern persists, I'll replace the CPU and retest. I'll let you know once finished. Thanks!
BSOD issues don't always point to the same cause as other problems might. If it matters, you might test the system without EXPO. Observe its behavior. Memory tests like memtest aren't perfect but can help confirm a theory. You could also remove unused components to see if the issue persists. Is the second NVMe drive required?
I've experienced crashes across various games, and they're spread between different drives. The latest game I played is on the second drive (2Tb), while I also had issues on the main drive (1Tb). It's important to note that the system mostly restarts rather than displaying a BSoD. Based on 15 crashes, I've encountered 2 BSoDs and 13 restarts. I've tested various RAM configurations, including Vengeance 2666MHz, but still faced crashes.
If RAM is excluded, not being used doesn’t guarantee it won’t affect performance quietly. We’re moving toward testing various solutions. It’s useful to power off devices entirely while checking if they’re linked to any reboots or failures.
I'll keep this in mind. I'll begin by reducing GPU usage and then check if my cable sleeve might be the problem. If not, I'll replace the CPU since I think it could be the cause. I recall a project where a motherboard was compatible with all Intel generations, but using a 14th-gen CPU caused issues. After swapping to a 13th-gen CPU, everything worked fine. The motherboard I received was replaced and functioned properly with the 14th-gen CPU. I suspect a similar situation might be happening, so I want to exclude the CPU from the faulty list. Thanks again!
In this scenario memory refers to the CPU cache. For the CPU it appears as accessible storage. WHEA doesn't track RAM directly; you'll find clarification in the official documentation. This segment doesn't specify the exact cause or location of the problem—it only indicates which hardware detected the issue. To identify the root cause, examine the RawData error packet in the event's Details section using the UEFI CPER guide. Navigate to C:\Windows\Minidump and verify the presence of any minidump files. If found, return to the Windows directory and transfer the folder to your Downloads folder (or use the desktop if OneDrive isn't available). Compress the file and attach it to a post. Remember, if no dump files exist, right-click and save them directly on the WHEA events page. A convenient method for viewing all WHEA events is applying a filter: in Event Viewer, go to Windows Logs → System, choose Filter Current Log, then select WHEA-logger from the Event Sources menu. Highlight any matching entries, right-click and save them. This consolidates all relevant logs into a single file.