Your screen goes blue mid-match and the error code vanishes before you can read it. Learning how to diagnose blue screen errors on a gaming pc starts with catching that code and matching it to the right test. Most crashes leave a trail in minidump files and Event Viewer logs.
The stop code tells you which hardware or driver path to investigate first.
Microsoft documents over 300 distinct bug check codes but fewer than twenty appear regularly on gaming rigs. A 2024 analysis of crash dumps from enthusiast forums showed VIDEO_TDR_FAILURE and WHEA_UNCORRECTABLE_ERROR accounting for nearly 40 percent of cases. The next step separates software faults from hardware instability.
How to Diagnose Blue Screen Errors on a Gaming Pc
Record the stop code and QR code on the blue screen. Open Event Viewer and check the System log for critical errors at the crash timestamp. Pull the latest minidump from C:\Windows\Minidump and load it into WinDbg to see the faulting module.
Run Windows Memory Diagnostic or MemTest86 to rule out RAM. Stress test CPU and GPU while monitoring temperatures and voltages.
Why the Blue Screen Happens During a Gaming Session
Gaming pushes every component harder than typical desktop loads. The CPU, GPU, memory, and power supply all hit sustained peaks that reveal marginal stability. A factory overclock, an aggressive XMP profile, or a voltage droop under load can trigger a stop code that never appears during web browsing.
Heat compounds the problem by shifting silicon thresholds.
Driver conflicts are another major cause. A new GPU driver may clash with an older chipset driver or with anti-cheat kernel modules. Windows updates sometimes replace a vendor driver with a generic one that lacks power management features.
The result is a crash that looks like hardware failure but resolves with a clean driver install.
Power delivery is the silent killer. A 750 W unit that tests fine at idle may sag on the 12 V rail when a 450 W GPU and a 250 W CPU spike together. That transient dip crashes the system without leaving a clear hardware fault code.
The stop code often points to the victim, usually the GPU or CPU, not the root cause.
Memory instability hides behind other symptoms. An XMP profile that passes a quick POST can fail after 20 minutes of thermal soak. The errors manifest as PAGE_FAULT_IN_NONPAGED_AREA or SYSTEM_THREAD_EXCEPTION_NOT_HANDLED with random instruction pointers.
Testing at stock JEDEC speeds isolates the profile as the variable.
First Decision Points: Stop Code, Boot State, and Recent Changes

Image source: Wikimedia Commons / Arriva436 (CC BY-SA)
Start with the stop code. VIDEO_TDR_FAILURE points to GPU timeout. WHEA_UNCORRECTABLE_ERROR signals a hardware bus or cache error.
KMODE_EXCEPTION_NOT_HANDLED usually means a kernel driver fault. CRITICAL_PROCESS_DIED suggests system file corruption or storage failure. Each code narrows the branch before you run a single test.
Next check boot state. If Windows loads normally, you can run diagnostics inside the OS. If the machine reboots before the desktop appears, you need Safe Mode or offline tools.
A system that won't POST at all moves the problem to hardware verification before any software step.
Identify the last change. A new GPU driver, a BIOS update, an XMP enable, a Windows cumulative update, or a hardware swap are the usual suspects. Roll back that single change first.
If the crashes started after a dust cleaning, reseat every cable and card. Physical disturbance causes more intermittent faults than component aging.
Use this decision table to pick your first test.
| Stop Code Family | First Test | Boot State Required |
|---|---|---|
| VIDEO_TDR_FAILURE | Clean GPU driver install with DDU | Normal or Safe Mode |
| WHEA_UNCORRECTABLE_ERROR | CPU stress test at stock clocks | Normal |
| KMODE_EXCEPTION_NOT_HANDLED | WinDbg minidump analysis | Normal |
| CRITICAL_PROCESS_DIED | SFC / DISM / chkdsk | Normal or Recovery |
| PAGE_FAULT_IN_NONPAGED_AREA | MemTest86 four passes | Offline USB |
| IRQL_NOT_LESS_OR_EQUAL | Driver verifier or rollback | Normal |
How to Diagnose the Branch: Minidumps, Event Logs, and Memory Tests
Configure Windows to create minidumps. Open System Properties, Advanced, Startup and Recovery, Settings. Select Small memory dump (256 KB) and confirm the dump folder is %SystemRoot%\Minidump.
This ensures every crash writes a file you can open later.
Open WinDbg as administrator. Load the latest .dmp file. Type !analyze -v and press Enter.
The output shows the bug check code, the faulting module name, and the stack trace. A module like nvlddmkm.sys or amdkmdag.sys implicates the GPU driver. A module like ntoskrnl.exe with varying offsets often points to memory or CPU.
Event Viewer adds context. Filter the System log for Level Critical and Error around the crash time. Look for Kernel-Power 41 (unclean shutdown), WHEA-Logger entries, or Disk surprises.
Application log may show game or anti-cheat crashes that preceded the bug check.
Memory testing needs more than the quick Windows pass. Run MemTest86 from a bootable USB for at least four full passes. Errors that appear only on pass three or four indicate thermal sensitivity.
Test each stick alone in the primary slot to isolate a bad module or a bad slot.
CPU stress testing reveals cooling and voltage issues. Run OCCT small data set for 30 minutes while watching core temperatures and package power. A crash or throttle under 90 °C suggests the cooler is mounted poorly or the voltage curve is unstable.
Prime95 blend test adds memory controller load.
GPU stress testing isolates the card. Run FurMark for 15 minutes at native resolution. Watch for artifacting, fan curve response, and power draw.
A crash here with VIDEO_TDR_FAILURE confirms GPU instability. If FurMark passes but games crash, the issue may be TDR timeout or a specific API path.
Recommended Actions: Drivers, Safe Mode, Hardware Isolation, BIOS

Image source: Openverse / Openverse contributor
Clean GPU driver installation removes corrupted files and stale registry keys. Download the latest driver from NVIDIA, AMD, or Intel. Run Display Driver Uninstaller in Safe Mode with the "Clean and restart" option.
Install the fresh driver without GeForce Experience or Adrenalin extras. This resolves most VIDEO_TDR_FAILURE and KMODE_EXCEPTION cases tied to display drivers.
Safe Mode loads only Microsoft-signed drivers. If the system is stable there, a third-party driver or startup service is the cause. Use msconfig to disable all non-Microsoft services and startup items.
Re-enable groups until the crash returns. This binary search isolates the culprit faster than guessing.
Hardware isolation follows the minimum viable POST principle. Remove all but one RAM stick, the boot drive, and the GPU. Test stability.
Add components back one at a time. A crash that returns with the second stick points to memory or slot. A crash that returns with a second drive points to storage or cable.
BIOS updates fix memory compatibility and CPU microcode bugs. Check the motherboard vendor support page for the latest version. Read the changelog for "AGESA update," "memory stability," or "PCIe stability." Flash from the BIOS USB tool, not from Windows.
Reset to defaults after flash, then reapply only XMP if needed.
Disable XMP or EXPO as a baseline test. Run at JEDEC speeds (DDR4 2666 or 3200, DDR5 4800 or 5200) with auto voltage. If crashes stop, the memory controller or the modules cannot sustain the rated profile.
You can often keep the frequency and loosen timings or raise VDDQ slightly.
Chipset drivers matter more than most builders assume. Install the latest AMD or Intel chipset package from the CPU vendor, not the motherboard CD. These drivers manage PCIe power states, USB controllers, and SMU firmware.
An outdated chipset driver can cause WHEA_UNCORRECTABLE_ERROR on Ryzen platforms.
Power supply testing under load requires a multimeter or a hardware monitor with 12 V rail reporting. Software readings are approximate. If the 12 V rail drops below 11.6 V during a FurMark plus Prime95 combined run, the unit cannot sustain transient spikes.
Replace with a unit rated 20 to 30 percent above peak system draw.
Mistakes to Avoid When Troubleshooting a BSOD
Reinstalling Windows before checking hardware wastes hours. A clean install on faulty RAM or a sagging PSU will crash again. Verify memory and power first.
The exception is CRITICAL_PROCESS_DIED with no hardware logs, then a repair install or clean install is the correct step.
Assuming the stop code name tells the whole story. VIDEO_TDR_FAILURE can come from a GPU overclock, a PSU dip, a driver bug, or a PCIe riser cable. The code is a category, not a diagnosis.
Always cross-reference with minidump analysis and stress tests.
Skipping Safe Mode testing. If you only test in normal mode, you cannot separate Windows drivers from third-party drivers. Safe Mode is the control environment.
Use it.
Running MemTest86 for one pass and declaring memory good. Errors often appear after thermal soak or on specific test patterns. Four passes minimum.
Eight for DDR5.
Ignoring the power supply because "it worked before." Capacitors age. A unit that handled a 300 W load may fail at 500 W after two years. Gaming loads are transient and brutal.
Test under combined CPU plus GPU stress.
Updating BIOS without reading the changelog. Some versions add support but regress memory training on certain kits. Check forums for your exact motherboard and RAM combination before flashing.
Overlooking cable reseats. A loose 24-pin, 8-pin EPS, or PCIe power cable creates intermittent contact that looks like a component failure. Reseat every power cable before condemning hardware.
Using driver verifier without a plan. Verifier adds overhead that can crash a stable system. Enable only for suspected drivers, reboot, reproduce, then disable.
Never leave it on for daily use.
Swapping parts blindly without a baseline. Test one variable at a time. Changing GPU, RAM, and PSU together teaches you nothing about which fixed it.
Trusting software temperature readings for VRM or VRAM. GPU hotspot and memory junction temps are often 15 to 20 °C higher than the core sensor. Use HWiNFO64 and watch the junction temperature field.
Final Decision Guide: Which Test to Run Next
If the stop code is VIDEO_TDR_FAILURE, clean the GPU driver with DDU in Safe Mode. That step resolves most gaming display crashes. If it persists, stress test the GPU with FurMark for fifteen minutes.
If the code is WHEA_UNCORRECTABLE_ERROR, drop to stock JEDEC memory speeds and disable XMP. Run OCCT small data set for thirty minutes. Crashes at stock point to the power supply or CPU voltage.
A multimeter check on the 12 V rail under load confirms or rules out the PSU. For guidance on swapping the power supply, follow the step-by-step guide.
If the code is KMODE_EXCEPTION_NOT_HANDLED, open the minidump in WinDbg. Microsoft documents the full list of bug check codes on its Learn site. The faulting module name reveals the driver to target.
Roll back or reinstall that driver. If the module is ntoskrnl.exe with no clear offset, memory is the suspect. Run MemTest86 for four passes from a bootable USB.
If the system boots but crashes only in games, the GPU driver or the game itself is the prime suspect. Update the driver and verify game files in Steam. A recent Windows cumulative update can also trigger this.
For a related issue, see freezes during gameplay.
If the machine will not boot past POST, hardware isolation comes first. Reseat every cable and card. Test with one RAM stick in the primary slot.
For adding more memory, the guide covers compatible speeds and timings. If you recently cleaned the case, reseat the GPU and RAM. For safely cleaning the case, follow the guide to avoid static damage.
When a single test does not resolve it, escalate to a clean Windows install on a known-good drive. Back up your data first. If the clean install still crashes, the hardware is faulty.
For severe cases, consider a full rebuild.
Frequently Asked Questions
What is the difference between a BSOD and a game crash?
A BSOD stops the entire operating system and displays a blue screen with a stop code. A game crash returns you to the desktop or closes the application. BSODs indicate a kernel-level fault involving drivers or hardware.
Game crashes happen in user space and come from the game engine or GPU driver. Check the Windows Event Viewer to confirm which type you experienced.
Can a bad power supply cause a blue screen?
Yes. A power supply that cannot sustain transient load spikes will cause voltage dips on the 12 V rail. The GPU or CPU may receive insufficient voltage during a gaming spike, triggering a stop code.
The error often appears as WHEA_UNCORRECTABLE_ERROR or VIDEO_TDR_FAILURE. Testing the PSU with a multimeter under combined CPU and GPU load is the definitive check.
How long does MemTest86 take to run?
A single pass on DDR4 takes about twenty to thirty minutes per gigabyte of installed RAM. Four passes, the recommended minimum, take roughly two to three hours for a typical 16 GB kit. DDR5 runs slightly slower due to higher density.
Run it overnight for a thorough check.
Should I update my BIOS to fix a BSOD?
Only if the motherboard vendor release notes specifically mention a memory stability or CPU microcode fix relevant to your issue. BIOS updates carry a small risk of bricking the board if the flash is interrupted. Back up your data and use the BIOS USB tool, not Windows.
If your system is stable, do not update for a BSOD alone.
Will reinstalling Windows fix a blue screen?
It can fix BSODs caused by corrupted system files, bad driver installs, or malware. It will not fix faulty RAM, a failing GPU, or an unstable power supply. Run hardware diagnostics first.
If those pass, a repair install or clean install is the correct next step.
1 thought on “How to Diagnose Blue Screen Errors on a Gaming PC”
Comments are closed.