Crash due to memory handling issues, not under heavy load, occurring after app closures and not during full idle state.
Crash due to memory handling issues, not under heavy load, occurring after app closures and not during full idle state.
OS - Windows 10 22H2 same, is it an OEM version (installed right away)?
Age of system: hardware 1 year, 7900x3d + 64gb, 5600 (2 sticks), 4090 motherboard, b650 steel legend, WiFi power supply, 1050w gold Enermax Revolution DFX.
Last week I encountered BSODs after BSOD attempts. Sometimes DLLs appear (sfc and dim report everything fine) and other times memory issues arise (e.g., Memory Management).
I ran memtest86 twice, all passed; vram tests also clean; GPU shows no visual problems (even under high load or while gaming). sfc and dim are working, I updated the PC, uninstalled old software, created a new user account.
Essentially, I’ve done almost everything except reinstalling everything because I didn’t want to reconfigure.
I’m unsure what’s happening. Bluescreens only occur during very light tasks—never in full idle or under heavy load (like gaming or using Handbrake). Occasionally, apps close unexpectedly.
Could it be a VDROOP? If reinstalling is the only option, can I upgrade to Windows 11 for similar behavior, or should a complete clean install be necessary?
012725-29562-01.dmp
temperatures seem acceptable overall, though the hdd reached 50°C just before the final BIOS update. That could be relevant. A user in another discussion noted my RAM was slightly elevated but within normal limits, and I added a fan aimed at those components afterward.
It could be a power unit problem. You might want to attempt a test run first. Avoid updating to version 11 as it requires significant old data. If your SSD is NVMe, the hardware may be faulty. Ensure your board has several NVMe ports; switch to another port and observe changes? Check if your BIOS is current.
Check if the voltages are within normal range, especially during idle. In BIOS you only enabled the EFI system control, which is fine.
The data in the dump files suggests memory issues, though RAM isn't always the clear answer. Windows moves low-priority RAM information into the page file and retrieves it when required, making storage appear as memory. The CPU’s memory controller plays a key role; failure here can mimic memory behavior. In about half of the dumps, storage or its drivers are blamed. Two entries involve storage drivers, but overall, storage isn’t the main culprit—though I won’t dismiss it entirely. If any component is overclocked or undervolted, remove it. Disable automatic overclocking features like Precision Boost Overdrive (PBO) and EXPO profiles on the RAM. The 5600MT/s speed is a minor overclock, but we avoid risky settings during testing. You’re running a relatively recent BIOS; updating it can help, especially if you upgrade to version 3.16. Remember to also refresh the Chipset driver if you install the latest BIOS, as this may resolve compatibility problems with your board. To verify RAM health, test each stick individually—use the second slot for CPU socket counting. If only one stick causes crashes, it’s likely a faulty stick. Memory testing tools often miss defective DDR4 and newer RAM, so manual checks are more reliable.
Part2 also asks if it's acceptable for the CPU to spike to 87°C during the test, especially with your 420AIO from an old Intel setup.