Full Diagnostic Tree & Step-by-Step Overview
What main symptom is your system exhibiting right now?
- Frequent Blue Screen crashes (BSOD) or unexpected system reboots
- PC turns on, fans spin, but NO DISPLAY (Black screen / Beep codes)
- Random game/app crashes to desktop with no clear error message
- File corruption, failing OS updates, or corrupted program installs
- System shows 'Usable RAM' much lower than total installed physical RAM
Are you seeing stop codes like MEMORY_MANAGEMENT, PAGE_FAULT_IN_NONPAGED_AREA, or IRQL_NOT_LESS_OR_EQUAL?
- Yes, one or more of these exact memory-related stop codes appear
- Stop codes vary constantly on every reboot (Random stop codes)
- Crashes only happen under heavy GPU load (e.g. 3D rendering / Gaming)
Does your motherboard show a DRAM Debug LED (Red/Yellow) or emit a specific beep code pattern at startup?
- Yes, DRAM Debug LED is lit solid or motherboard beeps continuously
- No debug LEDs, but system power loops (turns on and off every few seconds)
Do crashes happen across multiple completely unrelated programs (Chrome, Games, Discord)?
- Yes, multiple distinct programs crash randomly without warning
- Only ONE specific program or game crashes repeatedly
Faulty RAM often alters data in transit before it hits the drive, corrupting files. Are file downloads failing integrity/CRC checks?
- Yes, archive files (ZIP/RAR) or installers fail CRC/hash checks routinely
- No, files save fine; OS just crashes during heavy multitasking
In Windows Task Manager > Performance > Memory, does it state a large chunk of RAM is 'Hardware Reserved' (e.g. 16GB installed, only 7.9GB usable)?
- Yes, half or more of my installed RAM is listed as 'Hardware Reserved'
- No, full RAM capacity is seen, but system is sluggish and paging heavily
When the crash occurs, does the entire screen freeze with buzzing audio, or does Windows immediately display a BSOD?
- Displays a standard Windows BSOD crash screen
- Screen freezes solid or goes instantly black while GPU fans rev to maximum
Can your computer currently boot into Windows reliably enough to run diagnostic software?
- Yes, I can access Windows desktop easily
- No, Windows crashes during startup or fails to boot completely
Which testing tool would you prefer to start with?
- Built-in tool (Windows Memory Diagnostic - Fast & No downloads)
- Gold Standard tool (MemTest86 via USB drive - Most thorough)
- Advanced user tool (HCI MemTest or Karhu inside Windows)
Do you have access to another working PC and a USB flash drive to create a bootable memory tester?
- Yes, I can create a bootable MemTest86 USB drive on another PC
- No, I need to troubleshoot physically using stick manipulation right now
Running Built-in Windows Memory Test
Solution:
Diagnostic Execution
1. Press Win + R, type mdsched.exe, and hit Enter.
2. Select Restart now and check for problems (recommended).
3. Upon reboot, the tool will launch automatically in a blue environment. Press F1 to open options.
4. Change test mix to Extended, set pass count to 2, and press F10 to apply.
Interpreting Results
Status - Hardware problems were detected: At least one RAM stick or DIMM slot is failing bit-flip logic tests. Proceed to physical isolation.No errors detected: If crashes persist, run MemTest86 for a deeper 4-pass test, as WMD can miss high-frequency or temperature-sensitive address errors.
The Gold-Standard Hardware Memory Test
Solution:
Step-by-Step Guide
1. Download MemTest86 Free (PassMark) on a working PC.
2. Insert a USB drive and use the included imageUSB tool to burn MemTest86.
3. Plug the USB into the target PC, turn it on, and spam F12, F11, or F8 to open the Boot Menu.
4. Select the UEFI USB drive to boot directly into MemTest86.
5. Allow all 13 individual algorithm tests across 4 full passes (this takes 2-4 hours).
Interpreting Errors
0 Errors: RAM physical hardware, timing gates, and address lines are 100% healthy.1 or more Errors (Red lines): Physical RAM failure confirmed. Even a single error in MemTest86 indicates unstable memory that will corrupt data. Proceed immediately to physical stick isolation.
Testing Under Windows Load Conditions
Solution:
Executing TestMem5 (TM5)
1. Download TestMem5 and extract it.
2. Download the Universal2 or anta777 Heavy5 configuration profile.
3. Right-click TM5.exe and select Run as Administrator.
4. Let the test run until it reaches 100% coverage across all threads.
Verification
If error counters increase above 0, your memory profile, XMP timing, or physical memory cells are unstable under heat/voltage stress.
Isolating Bad Stick vs. Damaged Motherboard Slot
Solution:
The Single-Stick Isolation Workflow
If software tests show errors or the PC won't POST, you must pinpoint the exact broken component.
1. Shut down PC and unplug main power cable.
2. Remove ALL RAM modules from the motherboard.
3. Take Stick A only and insert it into DIMM Slot A2 (typically the 2nd slot away from the CPU socket).
4. Push down until both retention clips click firmly into place.
5. Power on PC and test for POST or run a memory test.
If Stick A works fine, set it aside.Test Stick B alone in Slot A2.If Stick B fails or prevents POST, Stick B is defective and needs replacement (RMA).6. Slot Testing: If a stick fails in Slot A2, test that same stick in Slot B2 to rule out a damaged motherboard DIMM slot or bent CPU socket pins.
Unstable Overclocking Profiles (XMP / EXPO / DOCP)
Solution:
Diagnostic Cause
Enabling XMP (Intel) or EXPO/DOCP (AMD) overclocks the integrated memory controller (IMC) on your CPU. Over time or with modern multi-kit configurations, these profiles can become unstable, simulating hardware failure.
Step-by-Step Fix
1. Disconnect wall power.
2. Remove the round CR2032 CMOS battery on motherboard for 5 minutes (or short the 2-pin CLR_CMOS jumper with a screwdriver for 10 seconds).
3. Reinsert battery and power on into BIOS (Del or F2).
4. Leave memory settings at DEFAULT JEDEC speed (e.g., 2133MHz, 2400MHz, or 4800MHz depending on DDR4/DDR5 standard) with XMP/EXPO turned OFF.
5. Retest system. If instability vanishes, the RAM sticks are healthy, but the high-frequency XMP/EXPO profile is unstable on your specific CPU memory controller.
Resolving Unmapped Memory Channel Allocation
Solution:
Diagnostic Cause
When Windows detects a stick electrically during POST but cannot establish stable dual-channel address maps, it marks the stick as 'Hardware Reserved'.
Step-by-Step Fix
1. Ensure RAM sticks are installed in the correct dual-channel slots (Slots A2 and B2 - 2nd and 4th slots from CPU).
2. Open Windows search, type msconfig, and open System Configuration.
3. Go to Boot tab > Advanced Options.
4. Ensure Maximum memory is UNCHECKED.
5. Apply changes, shut down, and reseat both RAM sticks securely.
Application-Specific Code Exception
Solution:
Diagnostic Conclusion
Because crashes are isolated to a single application while all other multitasking and stress tests pass, your physical RAM is healthy.
Recommended Steps
1. Reinstall the crashing application.
2. Clean-install display drivers using DDU (Display Driver Uninstaller).
3. Run System File Checker: Open CMD as Administrator and execute sfc /scannow.
Non-RAM Hardware Failure Identified
Solution:
Diagnostic Conclusion
Full screen freezes, display output cuts, or instant power shutoffs during heavy load point toward VRAM (Video RAM on graphics card), GPU thermal throttling, or PSU transient power spikes—not physical system RAM.
Recommended Steps
1. Monitor GPU temperatures using HWInfo64 or MSI Afterburner.
2. Ensure separate PCIe power cables are running from PSU to GPU (avoid daisy-chain pigtails).