I'm at my wit's end with repeated BSODs and critical errors.
I'm at my wit's end with repeated BSODs and critical errors.
Components listed include the Asus Rog Strix RTX 4090, i9 13900K, Asus ROG Strix Z790-F Gaming Wifi Motherboard, Kingston Fury Beast RGB Expo 64GB 6000MT/s DDR5, and Corsair RM1000X Power Supply. The WD Black sn850x 1TB is also mentioned. Issues reported involve kernel power errors, a PAGE_FAULT_IN_NONPAGED_AREA message, disabled desktop composition, and intermittent shutdowns during BSODs. The PC has been in use for about two to three years, with these problems emerging over the past three months. The main concern is the RAM, possibly due to its suboptimal fit for the build, though current pricing makes it hard to find alternatives. Various troubleshooting steps have been attempted, such as running Memtest, updating BIOS, reseating RAM, adjusting RAM speed, lowering GPU performance, and sending dump files to ChatGPT. Other changes include driver updates, Windows updates, and underpowering the GPU by 20%. The situation is causing concern, and a quick resolution is desired.
From the symptoms, a straightforward solution isn't available. The problem might stem from either software or hardware concerns, possibly both at the same time. From a software standpoint, the simplest remedy is to reformat the OS drive and perform a clean Windows installation. This should eliminate software-related issues, bloatware, and malware (except for firmware rootkits). If the system runs smoothly after a fresh OS install, that's ideal. However, if the issue persists and Kernel ID41 remains, then:
* Remove your GPU and connect the monitor to MoBo.
Check whether Kernel ID41 disappears. If it does, the problem likely lies with your GPU or power supply unit.
Also note, running an Intel 13th generation processor can cause performance drops due to aging; the fix might be replacing the CPU. But it's worth clearing the OS first (free) and also removing the GPU to further narrow down the cause.
A 1000W PSU is insufficient for your setup.
The RTX 4090 can handle up to 550W, while the i9-13900K supports around 413W; combined they use about 960W, leaving little headroom for other components.
A 1200W PSU would be the minimum requirement, and using a 1600W unit would provide more headroom.
Welcome to the forums, newcomer! Could you share the .dmp files so we can examine them and identify the cause? It would be helpful to know if you moved to version 0x12F before the problems started or if you joined later. This might help us determine if your processor is affected. Also, please include the BIOS version you're using for your motherboard. Updated all drivers, please—ideally without needing a third-party app?
It's uncommon to extract 400W from a CPU and 500W from a GPU. A realistic scenario would involve a 600-700W total load.
Rare maybe, but when such a rare event happens, PSU will cut power to the whole PC. This could cause software issues or even hardware failure. Personally, I wouldn’t take that risk. The RTX 4090 and i9-13900K aren’t inexpensive to replace. It would be much cheaper to purchase a proper PSU with at least 1.2kW or 1.6kW capacity. Safer to be cautious.
As rare as it sounds, achieving this would require significant artificial strain on both parts. For instance, in gaming, the 13900K doesn’t reach 200W under normal conditions.
It has been mentioned that you mentioned Event Viewer. Please also check the Reliability History/Monitor section. It offers a more user-friendly interface and the timelime format could help identify trends over time.
Additionally, ensure all critical data is backed up at least twice to separate locations from the host system. Confirm the backups can be recovered and read properly.
Power down the device, unplug it, and open the case. Remove dust and debris. Inspect all connectors, cards, RAM, jumpers, and case connections to confirm they are securely installed. Use a bright flashlight to look for any signs of damage such as exposed conductors, bent wires, melting, corrosion, or loose components.
It seems something may have become slightly loose over time due to heat expansion/contraction or vibrations.
You’ve been investigating the PSU and found it clearly failed. Yesterday I had to disconnect it to get anything working, and now it’s completely nonfunctional except for RGB from the motherboard.
Earlier tests showed:
- Memtest: 40 errors in 4 passes with 13 tests.
- Intel diagnostic tool: Passed, though the PC would occasionally freeze but no BSOD occurred.
- BIOS version: Updated about 2-3 months ago; prior updates were annual.
- Drivers: All updates came from official sources, no third-party apps.
I have my old dump files ready, but there’s no way to attach them directly in this message. Would you like me to suggest an alternative way to share them?
As mentioned earlier, the system was too weak for your setup. It would be better to consider an updated model with adequate capacity. The top picks are:
Best 1.2kW PSUs:
https://hwbusters.com/best_picks/best-at...busters/6/
Best 1.3+ kW PSUs:
https://hwbusters.com/best_picks/best-at...busters/7/
Hope the replacement PSU works well and keeps the rest of your system functioning smoothly. Given the quality of your build, it's unlikely the new PSU caused any issues. Still, no component is completely safe—when a problem arises, the motherboard will be the first to be affected, followed by the GPU. CPU and RAM are more robust but not entirely immune.
If you follow what the original poster said, you'll see they don’t use their PC for gaming; instead, it serves as a workstation, where workloads are generally heavier than in games.
Do you think it’s reasonable to suspect the RAM might be the problem given the memtest failure? I’ve attempted to boot with one stick at a time, successfully getting it to start with both sticks separately, but I still had to perform the unplugging trick repeatedly, which sometimes needed several tries. I’m just trying to consider all potential causes.