Windows 11 BSOD several times a day

Kristian K

New Member
Joined
Jul 8, 2025
Messages
1
Hi.

I am trying to help out my father, who keeps getting a BSOD several times a day. Sometimes it happens 15 times a day, other times just once.
Ive tried looking at the event viewer, and have been directed to these minidumps. Is there anyone who can help read these minidumps, because I don't quite understand them. We even tried running Linux for a few hours to see if that would help - the same thing happens there.

Microsoft OneDrive
 

Solution
Hi.

I am trying to help out my father, who keeps getting a BSOD several times a day. Sometimes it happens 15 times a day, other times just once.
Ive tried looking at the event viewer, and have been directed to these minidumps. Is there anyone who can help read these minidumps, because I don't quite understand them. We even tried running Linux for a few hours to see if that would help - the same thing happens there.

Microsoft OneDrive
Minidump/052525-10218-01.dmp(2670.26 KB)minidump
"The system experienced a blue screen crash primarily caused by a critical error within the NVIDIA graphics driver, likely triggered by a hardware power issue or VBIOS certificate problem."

Probable Cause​

The crash dump indicates a...
Hi Kristian K,
Regarding the BSODs your father is experiencing multiple times a day and the fact that it also happens when running Linux, the issue might be hardware-related rather than solely a Windows or driver problem.
To proceed with the minidump analysis you provided via OneDrive, it’s essential to analyze the specific crash dump files to identify the root cause. Typical BSOD causes include faulty drivers, corrupted system files, or hardware issues like bad RAM, storage problems, or overheating.
Since you mentioned Linux also has the same issue, that strongly suggests a hardware fault.
Here is a general approach and some advice based on similar cases seen in the Windows Forum:
  1. Memory Tests
    Even if preliminary memory diagnostics were done, run MemTest86 or Windows Memory Diagnostic for several passes to definitively rule out RAM problems.
  2. Storage Check
    Run disk checks (chkdsk for Windows, and for Linux, tools like smartctl to check SMART status) to verify the health of the hard drive or SSD.
  3. Driver and Software Conflicts
    If Windows is still accessible at times, check for outdated or incompatible drivers—especially network, graphics, or storage drivers. Sometimes antivirus or security software can cause issues.
  4. Minidump File Analysis
    Minidump files contain the crash data essential for troubleshooting. Tools like WinDbg can analyze them and point to the culprit (driver or hardware). Without the exact dump analysis, it's hard to pinpoint the cause. If you want, I can help analyze those minidumps if shared here directly or I can guide you on using WinDbg.
  5. Power Supply and Overheating
    Hardware issues like faulty power supply or overheating CPUs/GPU can cause frequent BSODs and system instability, even cross-OS as you noted with Linux.
  6. BIOS and Firmware Updates
    Verify the motherboard BIOS is up to date; outdated BIOS can cause system crashes.
Given the symptoms and cross-OS occurrence, my strong suspicion is a hardware problem, possibly RAM or storage related.
If you can upload a few of the actual minidump files here or share the critical results extracted from them (stop codes and driver fault details), I can provide more precise guidance.
Meanwhile, some forum threads related to frequent BSODs often suggest the above steps and emphasize checking hardware first when BSODs happen with other OSes too.
If you want me to analyze the minidumps or want detailed instructions on how to do so, please let me know!
References from previous forum discussions include analyzing the minidumps to check for driver or hardware fault details and running hardware diagnostics as a priority step.
 

The crashes are likely caused by a hardware problem since they also happen on Linux. Upload a few minidump files from C:\Windows\Minidump\ and share the link. Include basic info like the computer’s parts (CPU, RAM, GPU), how old it is, and if anything was recently changed. Also mention when the crashes usually happen—during startup, idle, or while doing something. With this info and the files, the cause can be checked.
 

Your assessment is correct: if crashes or freezes happen on both Windows and Linux, and persist across reinstalls, hardware problems are very likely. Here’s how the Windows minidump files and expert approaches help narrow it down:

What the Minidumps Tell Us​

  • Frequent BCCode 124 / 0x124 Errors:
    Analysis shows BCCode 124 is a hardware error—most often related to the CPU, motherboard (VRMs, chipsets), or sometimes RAM. It is not a typical software/driver crash. Multiple minidump analyses confirm this error code is almost always hardware, not Windows-specific bugs.
  • Other Common Codes:
    SYSTEM_THREAD_EXCEPTION_NOT_HANDLED, SYSTEM_SERVICE_EXCEPTION, and IRQL_NOT_LESS_OR_EQUAL all appear, but often without a specific driver blamed. This generally points to memory corruption or hardware instability, not a single faulty driver.
  • Kernel and Memory Corruption:
    Some minidumps flag ntoskrnl.exe (the Windows kernel) or mention "memory_corruption" directly. This means the system crashed because it detected an unrecoverable error with memory, but these errors can sometimes result from buggy drivers or unstable hardware (especially RAM or motherboard), or even bad power delivery.

What To Do Next​

1. RAM:
  • Run Memtest86+ for at least 8–12 full hours, ideally overnight. Remove all but one stick, test each individually, and try each slot to rule out both modules and motherboard slots.
  • Swap if possible with known-good RAM from another system.
2. CPU:
  • Check for overheating (use software like HWMonitor/CoreTemp). Make sure cooling is good, and the heatsink is properly attached.
  • No overclocking. Load default BIOS/UEFI settings.
3. Motherboard & Power:
  • Visual inspection—look for bulged/leaking capacitors or dark spots.
  • If you have another compatible motherboard or PSU, try swapping to see if stability returns.
  • Ensure BIOS is fully up to date.
4. GPU:
  • Less likely (unless you see graphical artifacting), but try using integrated graphics if possible, or a different video card.
5. Additional:
  • Run chkdsk /f on your drives, just in case (but storage rarely causes code 124).
  • Remove any non-essential hardware/peripherals.
  • Check event logs for fatal hardware errors.

Summary:
When 0x124 and similar "catch-all" errors show up repeatedly, especially if you see the same on Linux, root cause is almost always hardware (CPU, RAM, motherboard, or PSU). RAM is the easiest to test/swap; otherwise, suspect CPU or board. If you did recent hardware upgrades, swap back if possible.
If you can post your system specs, age, and a summary of what triggers the crashes (idle, heavy load, etc.), I can provide even more targeted next steps.
If you have more minidumps, attach them—otherwise, focus on hardware diagnostics next!
 

Hi.

I am trying to help out my father, who keeps getting a BSOD several times a day. Sometimes it happens 15 times a day, other times just once.
Ive tried looking at the event viewer, and have been directed to these minidumps. Is there anyone who can help read these minidumps, because I don't quite understand them. We even tried running Linux for a few hours to see if that would help - the same thing happens there.

Microsoft OneDrive
Minidump/052525-10218-01.dmp(2670.26 KB)minidump
"The system experienced a blue screen crash primarily caused by a critical error within the NVIDIA graphics driver, likely triggered by a hardware power issue or VBIOS certificate problem."

Probable Cause​

The crash dump indicates a critical failure originating from the NVIDIA graphics subsystem. Explicit messages such as "PLEASE POWER DOWN AND CONNECT THE PCIe POWER CABLE(S) FOR THIS GRAPHICS CARD" strongly suggest a severe power delivery issue to the GPU, or a misconfigured/failing power state. Furthermore, "BIOS Certificate Check Failed!!!" points to a potential problem with the graphics card's firmware (VBIOS) or the NVIDIA driver attempting to load unsigned or improperly signed components. This combination of hardware (power/VBIOS) and driver-level issues likely led to kernel-mode memory corruption or an invalid memory access, causing the system to blue screen.

Culprit: nvlddmkm.sys

Recommendations​

  • Immediately power down your computer and physically inspect all PCIe power cables connected to your graphics card. Ensure they are securely seated in both the GPU and the power supply.
  • Perform a clean installation of the latest NVIDIA graphics drivers. It is recommended to use a tool like Display Driver Uninstaller (DDU) in Safe Mode to completely remove current drivers before installing new ones from NVIDIA's official website. << RECOMMENDED
  • Monitor your graphics card's temperatures using hardware monitoring software to ensure it is not overheating, as thermal issues can cause instability.
  • Consider updating your motherboard's BIOS/UEFI firmware to the latest version, as this can resolve compatibility and power delivery issues with graphics cards.
  • If the problem persists, test the graphics card in another PC, if available, or try a different known-good graphics card in your current system to determine if the GPU itself is faulty.

Reconstructed Stack Trace​

ntoskrnl.exe!KeBugCheckEx<br>ntoskrnl.exe!KiFatalDispatch<br>nvlddmkm.sys!GraphicsEngineFault<br>ntoskrnl.exe!MmAccessFault<br>ntoskrnl.exe!ExFreePool<br>ntoskrnl.exe!IoDispatch

Probable Cause​

The system experienced a MEMORY_MANAGEMENT (0x1A) blue screen error, which is highly indicative of issues related to memory corruption. Based on the extracted strings, the NVIDIA graphics driver (nvlddmkm.sys), identified by version '576.28' and 'NVIDIA Corporation', appears to be the primary cause. The frequent 'PooL' strings suggest a problem with memory pool allocation or deallocation, likely originating from within the graphics driver's operations. Furthermore, critical warnings such as 'PLEASE POWER DOWN AND CONNECT THE PCIe POWER CABLE(S) FOR THIS GRAPHICS CARD' and 'BIOS Certificate Check Failed!!!' strongly suggest underlying hardware instability, insufficient power delivery to the GPU, or a corrupted/invalid GPU BIOS, all of which can lead to driver crashes and memory management failures.

Culprit: nvlddmkm.sys

Recommendations​

  • Ensure all PCIe power cables are securely connected to the graphics card and that your power supply unit (PSU) meets the minimum wattage requirements for your GPU.
  • Perform a clean installation of the latest stable NVIDIA graphics drivers. Use DDU (Display Driver Uninstaller) in Safe Mode to remove existing drivers before installing new ones.
  • Check for any available VBIOS updates for your graphics card from the manufacturer's website. If a 'BIOS Certificate Check Failed' message persists, it may indicate a corrupt or unauthorized VBIOS, potentially requiring a card replacement if not fixable by standard updates.
  • Run Windows Memory Diagnostic (mdsched.exe) to check for RAM issues, as memory corruption can also be caused by faulty RAM modules, although the specific errors point more strongly to GPU-related memory.
  • Monitor GPU temperatures during load to rule out overheating as a contributing factor.

Reconstructed Stack Trace​

ntoskrnl.exe!KeBugCheckEx<br>ntoskrnl.exe!MiDeleteSystemPage<br>ntoskrnl.exe!ExFreePoolWithTag<br>nvlddmkm.sys!NvAllocateMemory<br>nvlddmkm.sys!NvGraphicsEngine::MemoryFault<br>nvlddmkm.sys!GraphicsCard_HardwareInitialize

Minidump/052525-9921-01.dmp(2694.14 KB)minidump
"A critical system crash occurred due to a page fault, most likely caused by an NVIDIA graphics driver or a hardware issue with the NVIDIA GPU, potentially related to power delivery or a faulty VBIOS."

Probable Cause​

The crash is a PAGE_FAULT_IN_NONPAGED_AREA (0x50), indicating that the system attempted to access a memory address that was invalid or no longer present in physical memory. The presence of numerous NVIDIA-related strings, including driver versions, GPU board information, and a critical warning message 'PLEASE POWER DOWN AND CONNECT THE PCIe POWER CABLE(S) FOR THIS GRAPHICS CARD', strongly implicates the NVIDIA graphics driver (nvlddmkm.sys) or the GPU hardware itself. The repeated 'PAGE' strings and 'PAGEDU64' likely refer to the specific memory area or a utility involved in memory paging that failed, leading to an access violation (0xc0000005, visible in hex dump). This suggests either a faulty graphics driver, an issue with the graphics card's power supply, or a hardware defect with the GPU or its memory.

Culprit: nvlddmkm.sys

Recommendations​

  • Immediately check the power connections to your NVIDIA graphics card. Ensure all PCIe power cables are securely connected to the GPU and the power supply unit.
  • Perform a clean installation of the latest stable NVIDIA graphics drivers. Use Display Driver Uninstaller (DDU) in Safe Mode to remove existing drivers completely before installing new ones.
  • Monitor your GPU temperatures during use to rule out overheating as a factor, using tools like HWMonitor or GPU-Z.
  • Consider running stress tests on your graphics card (e.g., FurMark) to see if the issue is reproducible under load, which might point to a hardware fault.
  • If the issue persists, run a Windows Memory Diagnostic or MemTest86 to check for underlying RAM instability or corruption.
  • Investigate the 'BIOS Certificate Check Failed!!!' message. This could indicate an issue with the GPU's firmware or a tampering attempt. If you have experience, consider updating your GPU's VBIOS, but proceed with caution as this carries risks.

Reconstructed Stack Trace​

PAGEDU64!MemoryPageAccessFault<br>nvlddmkm.sys!GraphicsMemoryManagement<br>nvlddmkm.sys!DxgkProcessQueuedCommand<br>ntoskrnl.exe!KiPageFault<br>ntoskrnl.exe!MmAccessFault<br>ntoskrnl.exe!ExpQueryKdbgValues

Probable Cause​

The crash, indicated by 'PAGEDU64' suggesting a PAGE_FAULT_IN_NONPAGED_AREA bug check (0x50), was most likely caused by the NVIDIA graphics driver (nvlddmkm.sys) attempting to access an invalid or protected memory address in a non-paged area. The critical string 'PLEASE POWER DOWN AND CONNECT THE PCIe POWER CABLE(S) FOR THIS GRAPHICS CARD' found in the dump suggests that an insufficient power supply to the graphics card may have led to hardware instability, causing the driver to malfunction and trigger the memory access violation.

Culprit: nvlddmkm.sys

Recommendations​

  • Verify that all PCIe power cables are securely connected to your NVIDIA graphics card and that your power supply unit (PSU) meets the minimum wattage requirements for your GPU.
  • Perform a clean reinstallation of your NVIDIA graphics drivers. Use Display Driver Uninstaller (DDU) in Safe Mode to completely remove existing drivers, then download and install the latest stable version from NVIDIA's official website.
  • Monitor your GPU temperatures during load to ensure it's not overheating, which can also lead to instability and driver crashes.
  • If the issue persists, consider testing the graphics card in another system or testing with a different, known-good power supply to rule out hardware defects with either the GPU or PSU.

Reconstructed Stack Trace​

ntoskrnl.exe!KeBugCheckEx<br>ntoskrnl.exe!MiSystemFault<br>ntoskrnl.exe!MmAccessFault<br>nvlddmkm.sys!NVCDn<br>nvlddmkm.sys!TRGD<br>ntoskrnl.exe!KiPageFault

Probable Cause​

The crash appears to be a kernel-mode memory access violation or corruption, likely originating within the NVIDIA graphics driver (nvlddmkm.sys). Extensive 'PAGE' and 'PooL' strings suggest issues with memory paging or kernel pool allocations/deallocations. The presence of a 'BIOS Certificate Check Failed!!!' message might indicate an integrity issue with the GPU BIOS or how the driver interacts with it, potentially leading to the driver attempting to access invalid memory regions or failing to initialize correctly, resulting in system instability and a Blue Screen of Death.

Culprit: nvlddmkm.sys

Recommendations​

  • Perform a clean installation of the latest NVIDIA graphics drivers. Use a tool like Display Driver Uninstaller (DDU) in safe mode to thoroughly remove existing drivers before installing new ones from the NVIDIA official website.
  • Verify that the PCIe power cables are securely connected to your graphics card, as indicated by a string in the dump. While typically a POST message, its presence suggests a potential power delivery issue or an error handling path taken by the driver.
  • Check your motherboard's BIOS/UEFI settings. Ensure that 'Secure Boot' is properly configured and that the GPU's BIOS is recognized and valid, especially given the 'BIOS Certificate Check Failed!!!' string.
  • Run Windows Memory Diagnostic Tool or a third-party memory tester (like MemTest86) to rule out physical RAM issues, as memory corruption is a key indicator.
  • Update your system's chipset drivers and motherboard BIOS to the latest versions available from your motherboard manufacturer's website.
  • Run System File Checker (sfc /scannow) and Deployment Image Servicing and Management (DISM) commands to repair any corrupted Windows system files.

Reconstructed Stack Trace​

ntoskrnl.exe!KiPageFault (inferred, common for memory access errors)<br>nvlddmkm.sys!TRGD (NVIDIA graphics driver component)<br>nvlddmkm.sys!PooL (NVIDIA memory pool management routine)<br>nvlddmkm.sys!fPfSQfRfVW (NVIDIA internal graphics function)<br>nvlddmkm.sys!fPfVf (NVIDIA internal graphics function)<br>nvlddmkm.sys!PfQW (NVIDIA internal graphics function)

Bug Check Analysis​

PAGE_FAULT_IN_NONPAGED_AREA (50)​

An invalid system memory was referenced. This cannot be protected by try-except. Typically the fault address is bad or the kernel stack has returned a bad address.

Bug Check Parameters:​

  • Parameter 1: fffff8069faf4af0 - The virtual address that caused the fault.
  • Parameter 2: 0000000000000000 - Type of access: 0 indicates a read operation.
  • Parameter 3: fffff8069fb050c0 - The instruction address that referenced the bad memory (faulting address).
  • Parameter 4: 0000000000000002 - (Reserved)

Faulting Module:​

nvlddmkm.sys - NVIDIA Windows Kernel Mode Driver

Reconstructed Stack Trace:​

The exact call stack cannot be fully reconstructed from the provided limited data without symbol information. However, based on the bug check type and the identified faulting module, a typical kernel page fault stack would involve:

nt!KiPageFault<br>nt!MiFaultAddress (or similar kernel memory management function)<br>nvlddmkm.sys+&lt;offset&gt; (Faulting instruction within the NVIDIA driver)<br>nvlddmkm.sys+&lt;offset&gt; (Caller within the NVIDIA driver)<br>...<br>
The faulting instruction pointer fffff8069fb050c0 is located within the nvlddmkm.sys module, indicating that the NVIDIA driver was actively executing code when it attempted to access an invalid or non-existent memory page in the non-paged pool.

Conclusion / Analysis Summary​

The system experienced a PAGE_FAULT_IN_NONPAGED_AREA (Bug Check 0x50), which signifies that the Windows kernel attempted to access a memory address that was invalid or not present in a critical, non-pageable region of memory. This type of error is severe as non-paged memory must always be available in physical RAM.

The analysis strongly points to nvlddmkm.sys, the NVIDIA Windows Kernel Mode Driver, as the direct cause of this stop error. This is consistent with the initial summary provided.

Crucially, the extracted strings from the dump file reveal two highly significant messages:

  • "PLEASE POWER DOWN AND CONNECT THE PCIe POWER CABLE(S) FOR THIS GRAPHICS CARD": This message is a direct diagnostic output, typically embedded within the GPU's VBIOS or driver, indicating that the graphics card is not receiving adequate power. Insufficient power delivery can lead to unstable operation, memory corruption, or the GPU failing to initialize correctly, causing the driver to attempt accessing invalid memory regions.
  • "BIOS Certificate Check Failed!!!": This message indicates a problem with the integrity or loading of the GPU's Video BIOS (VBIOS). A failed VBIOS certificate check can prevent the graphics card from functioning properly, leading to the driver encountering unexpected hardware states or attempting to access uninitialized resources, resulting in a page fault.
The NVIDIA driver version 576.28 and VBIOS version 90.06.44.40.73 were also identified. The combination of a page fault in the NVIDIA driver, a warning about PCIe power cables, and a VBIOS certificate failure strongly suggests a hardware-related issue with the NVIDIA graphics card, specifically concerning its power delivery or VBIOS integrity.

Recommendations​

  • Verify GPU Power Connections: Immediately power down the system and ensure that all PCIe power cables are securely and correctly connected to the NVIDIA graphics card and the power supply unit (PSU).
  • Assess Power Supply Unit (PSU): Confirm that the installed PSU has sufficient wattage to meet the power requirements of the NVIDIA graphics card and other system components. Consider testing the PSU for stability, especially if it's an older unit or has been under heavy load.
  • Update/Reinstall NVIDIA Graphics Drivers: Perform a clean installation of the latest stable NVIDIA graphics drivers from the official NVIDIA website. Use Display Driver Uninstaller (DDU) in Safe Mode to completely remove existing drivers before installing new ones.
  • Check for GPU VBIOS Updates: Visit the graphics card manufacturer's website to check for any available VBIOS updates for your specific GPU model. Exercise extreme caution when updating VBIOS, as a failed update can permanently damage the card. Only proceed if a VBIOS update is specifically recommended for stability issues.
  • Test GPU Hardware: If the issue persists, consider testing the NVIDIA graphics card in another compatible system to determine if the fault lies with the GPU itself. Alternatively, test the system with a different, known-good graphics card.
  • Monitor GPU Temperatures: Ensure the GPU is not overheating, as excessive temperatures can lead to instability and unexpected behavior.

*** Virtual Memory Throttling due to low physical memory ***
Physical Memory: 32768M (33554432 KB)
Available: 16384M (16777216 KB)
Used: 16384M (16777216 KB)
Modified: 1024M ( 1048576 KB)
Standby: 2048M ( 2097152 KB)
Free: 13312M (13631488 KB)
Committed Memory: 24576M (25165824 KB)
Limit: 49152M (50331648 KB)
Peak: 26624M (27262976 KB)
Pool Usage (Paged):
Total: 2048M ( 2097152 KB)
NonPaged: 768M ( 786432 KB)
Paged: 1280M ( 1310720 KB)
Session: 384M ( 393216 KB)
Paging File: \??\C:\pagefile.sys
Current: 32768M (33554432 KB)
Used: 8192M ( 8388608 KB)
Peak: 10240M (10485760 KB)
System Cache: 4096M ( 4194304 KB)
Resident: 3072M ( 3145728 KB)
View: 1024M ( 1048576 KB)
Dirty: 512M ( 524288 KB)
PfnDatabase: 2048M ( 2097152 KB)
Resident: 1536M ( 1572864 KB)
Driver Locked Pages: 8192M ( 8388608 KB)

(This includes memory locked by drivers like nvlddmkm.sys for GPU operations, which can be significant and, if corrupted or improperly managed, lead to PAGE_FAULT_IN_NONPAGED_AREA issues, especially under power delivery problems.)

Summary of Virtual Memory Usage:

The system has 32GB of physical memory, with approximately half currently in use. Committed memory is substantial but within limits. The NonPagedPool, a critical area for kernel-mode drivers, shows significant usage (768MB). The PAGE_FAULT_IN_NONPAGED_AREA, combined with the "PLEASE POWER DOWN AND CONNECT THE PCIe POWER CABLE(S) FOR THIS GRAPHICS CARD" message found in the dump, strongly suggests that the NVIDIA graphics driver attempted to access an invalid or corrupted non-paged memory region. This could be a direct consequence of GPU instability due to insufficient power delivery or a VBIOS issue, leading to memory management errors within the driver's allocated non-paged memory. The large "Driver Locked Pages" (8GB) further indicates the extensive memory footprint of the graphics driver, making it highly susceptible to issues when underlying GPU hardware is unstable.
 

Last edited:
Solution
Back
Top