01.How to Approach Diagnosis — Methodology
Efficient hardware diagnostics follow a basic rule: isolate one component at a time. Trying to test everything at once only creates confusion. The correct methodology is:
Random blue screen = likely RAM or driver. PC shuts down under load = likely overheating or weak power supply. PC doesn't turn on = power supply, motherboard, or RAM.
Always test software before hardware. A corrupted driver causes the same symptoms as bad RAM. Clean drivers → test RAM → test SSD → test CPU/GPU → test PSU.
Take screenshots of test results. If you go to a repair shop later, this saves them hours of work (and you money).
⚠️ Before Anything: Check Temperatures
Download HWiNFO64 and monitor temperatures while reproducing the problem. A PC that shuts down at 95°C+ on the CPU or GPU doesn't need complex diagnostics — it needs thermal paste replacement or better cooling. Check our temperature monitoring guide.
Don't do it Manually.
Voltris Optimizer automates this entire guide and removes Windows delay in seconds.
02.RAM Test — Suspect Number 1
RAM is responsible for more than 40% of random blue screen cases in apparently healthy computers. The problem is that RAM errors are hard to reproduce — they appear at random times.
🛠️ Tool #1: TestMem5 (Windows, 30 minutes)
The best RAM test that runs within Windows itself.
- Download TestMem5 (search Google for "TestMem5 anta777 download").
- Also download the extreme4.cfg profile by author anta777.
- Place the .cfg file in the same folder as tm5.exe.
- Run tm5.exe as Administrator.
- Let it run for at least 2 complete cycles (1-2 hours).
- Any red error = bad RAM.
🛠️ Tool #2: MemTest86+ (Before Boot)
More comprehensive as it runs outside Windows and tests pure RAM.
- Download MemTest86+ at memtest.org.
- Use Rufus to create a bootable USB drive.
- Restart the PC and boot from the USB drive.
- Let it run for at least 8 passes (2-4 hours).
- Zero errors = healthy RAM.
💡 Tip: XMP/EXPO Profile
If you have XMP/EXPO enabled in the BIOS and the RAM fails testing, try disabling XMP and testing at base frequency (2133/3200 MHz). Unstable RAM at XMP doesn't necessarily mean defective RAM — it could be that the processor's IMC doesn't support that specific frequency.
03.SSD and HDD Test — CrystalDiskInfo
Failing disks cause file corruption, Windows crashes, extreme slowdowns, and in critical cases, the PC might not even boot. S.M.A.R.T. technology (Self-Monitoring, Analysis, and Reporting Technology) allows detecting imminent failures before the disk dies.
How to use CrystalDiskInfo:
- Download and install CrystalDiskInfo (crystalmark.info — free version).
- Open the program. It will list all system disks.
- Check the Health Status at the top of each disk:
The disk is healthy. Continue using normally but maintain regular backups.
Backup NOW. The disk is deteriorating. Buy a replacement and migrate your data.
STOP using the computer. The disk could die at any moment. Immediate emergency backup required.
Critical S.M.A.R.T. Attributes to monitor:
| Attribute | Interpretation | Alarm |
|---|---|---|
| 05 | Reallocated Sectors (replaced bad sectors) | Any value > 0 |
| C5 | Current Pending Sectors (sectors waiting for reallocation) | Any value > 0 |
| C6 | Uncorrectable Sectors (irrecoverable sectors) | Any value > 0 |
| E9 | NAND Endurance (SSD lifespan) | Below 10% |
04.CPU and GPU Testing — Stress Test with OCCT
An unstable processor or graphics card might not show errors in normal use, but will crash or shut down during gaming or rendering. OCCT is the community's most respected tool for these tests.
Downloading and configuring OCCT:
- Download OCCT at ocbase.com (free version is sufficient).
- Install and open the program.
- Before starting any test, ensure the temperature monitor is visible (graph icon on the sidebar).
🖥️ CPU Test (Linpack)
Select "CPU: Linpack" in OCCT. Duration: 15-30 minutes.
Trouble sign: Temperature above 95°C, PC shuts down or restarts during testing, or OCCT reports a calculation error.
🎮 GPU Test (3D)
Select "GPU: 3D" or use FurMark separately. Duration: 15-20 minutes.
Trouble sign: Visual artifacts (random colored pixels), driver crashing ("Display driver stopped responding"), or PC shutting down.
⚡ Power Supply Test
Select "Power Supply" in OCCT. This will stress CPU and GPU simultaneously. Duration: 15-30 min.
Trouble sign: PC shuts down (PSU can't handle peak load), restarts, or OCCT detects a voltage drop on 12V/5V rails.
Important Tip: If the PC shuts down under the Power Supply test but handles CPU and GPU separately, the problem is almost certainly the power supply being insufficient for the configuration. Check if the PSU has adequate wattage with the be quiet! PSU calculator.
05.Interpreting Blue Screens (BSOD) — Reading the Code
Every blue screen has an error code. Knowing what the code is saves hours of unnecessary diagnostics. Note or photograph the code before the PC restarts.
MEMORY_MANAGEMENT (0x0000001A)
Almost always faulty RAM. Test with TestMem5 and MemTest86+.
PAGE_FAULT_IN_NONPAGED_AREA (0x00000050)
Bad RAM or corrupted driver. Test RAM and reinstall video drivers with DDU.
DPC_WATCHDOG_VIOLATION (0x00000133)
Outdated driver or firmware. Update SSD drivers and firmware.
DRIVER_IRQL_NOT_LESS_OR_EQUAL
Problematic driver. Use DDU to remove and reinstall the video driver or any other recently installed driver.
For in-depth BSOD diagnostics, visit our BSOD Dump Analysis Guide where we teach how to open .dmp files and identify exactly which driver file caused the blue screen.
Don't do it Manually.
Voltris Optimizer automates this entire guide and removes Windows delay in seconds.
Written by a verified expert
Douglas Felipe M. Gonçalves
Expert in Windows system optimization with years of experience in hardware diagnostics, kernel tuning, and advanced technical support. Founder of Voltris and developer of the Voltris Optimizer.
Meet the Voltris TeamConclusion and Next Steps
By following this guide on Hardware Diagnostic: How to Find Out Which PC Part Is Failing (2026), you are equipped with the verified technical knowledge to solve this issue with confidence.
If you still have difficulties after following all steps, our expert support team is available for a personalized remote diagnosis. Every system is unique and may require a specific approach.