My roughly one-year-old PC has started suffering intermittent hard crashes. The display suddenly loses signal, the fans—especially the GPU fans—run at full speed, and the system becomes completely unresponsive until I hold the power button. There is no blue screen, and it can happen while gaming, sitting idle, at the Windows login screen, or when waking from sleep. The crashes sometimes occur in clusters: the system may be stable for days or weeks, then crash repeatedly within a short period.
System specs:
- AMD Ryzen 7 7800X3D
- ASUS B650M-AYW WIFI motherboard
- Gigabyte RTX 5070 Eagle OC 12GB
- 16GB Kingston DDR5-6000, running at JEDEC defaults of 4800 MT/s with EXPO disabled
- MSI MAG A850GL 850W 80+ Gold PSU
- Windows 11 Home 25H2
Event Viewer only reports Kernel-Power Event ID 41, with BugcheckCode 0 and all parameters set to zero. I reset the CMOS, updated the BIOS, disabled PBO, checked that SoC voltage is on Auto with no offset, and confirmed the memory is not using EXPO. HWiNFO logging showed the 12V rail staying around 12.172V and temperatures rising normally around 45–48°C during one incident, although the crash itself may prevent the final readings from being recorded.
I also performed a clean Windows installation, removed Armoury Crate and similar software, and reseated the RAM and GPU power cables. The Minidump folder is empty. What hardware or settings should I test next?
4 Answers
The symptoms could fit an unstable CPU, even when the BIOS is set to stock settings. Some processors become unstable under particular boost or voltage conditions without producing a blue screen. As a diagnostic—not necessarily a permanent fix—try lowering the CPU boost limit or applying a small negative curve offset, then see whether the crashes stop. If reducing boost or voltage makes the problem disappear, that points toward CPU, motherboard power delivery, or firmware behavior.
This looks more like a hard hardware or firmware lockup than a normal Windows crash. Event 41 is only reporting that the machine was forced off; with BugcheckCode 0 and no minidump, Windows never got a chance to record the actual failure. A black screen followed by fans staying at 100% can happen when the CPU or GPU stops responding.
I would test methodically: run with one RAM stick at a time, swap slots, test the GPU if you have another card available, and try the system with the discrete GPU removed if the motherboard supports display output from the CPU. Also test another known-good PSU or power cable if possible. A stable 12V reading in software does not completely rule out transient power problems.
Because there are no dump files, checking the Minidump folder will not provide much useful information unless Windows is configured to create them for bugchecks. The next useful step is isolation rather than more reinstalls: run memory tests, try each DIMM separately, load BIOS defaults, test without sleep enabled, and check whether the crash occurs with the GPU removed or replaced. Keep a record of which configuration survives the longest, since the clustered nature of the failures can make a single successful test misleading.
I had a similar failure where the display cut out and the fans stayed at full speed until I held the power button. In my case it was the GPU driver or GPU path, so I would try a clean graphics-driver removal and reinstall, then test with a different driver version. If the issue still happens outside games, during login, or while waking from sleep, hardware is still more likely than a normal game or application problem.
That matches mine exactly: the screen loses signal and the GPU fans definitely ramp up and remain there until I shut the system down manually.