Test memory and stability

Decide whether memory is a relevant hypothesis

Memory testing is relevant when crashes occur under RAM-heavy workloads, stop codes vary, file decompression or compilation produces inconsistent errors, or instability began after memory installation or firmware tuning. A single slow application is not enough by itself.

Microsoft's blue-screen guidance includes Windows Memory Diagnostic as one advanced check and recommends correlating results with Event Viewer and other crash evidence. The tool is part of a wider process, not a standalone root-cause engine.

Prepare for the restart and run the built-in diagnostic

Microsoft documents launching Windows Memory Diagnostic with mdsched and choosing 'Restart now and check for problems.' The test runs before Windows loads and restarts the device when complete.

Save all work, close applications and disconnect unnecessary external devices. Return firmware memory settings to vendor defaults before interpreting instability caused by overclocking or custom timings, but do not change BIOS settings unless you understand the device-specific process.

1. 1

   Back up current work and close every application.

2. 2

   Press Windows + R, type mdsched.exe and press Enter.

3. 3

   Choose Restart now and check for problems.

4. 4

   Allow the standard test to complete without forcing power off.

5. 5

   Sign in after Windows restarts.

6. 6

   Wait for the notification or open Event Viewer to locate the result.

Find and preserve the result

If the notification disappears, open Event Viewer and search the System log for MemoryDiagnostics-Results around the test time. Record the event level, message and timestamp. Do not rely on memory of a brief notification.

A detected problem is strong evidence of instability but does not alone identify which module, slot, controller setting or motherboard condition caused it. A clean result means this pass did not detect an error; intermittent faults can remain.

1. 1

   Open Event Viewer after sign-in.

2. 2

   Select Windows Logs > System.

3. 3

   Filter or search for MemoryDiagnostics-Results.

4. 4

   Open the event matching the test timestamp.

5. 5

   Copy the result text into the troubleshooting record.

6. 6

   Keep the crash and workload evidence beside the diagnostic result.

Interpret the result with configuration and workload context

If errors are detected, stop memory-intensive work that could corrupt files and verify backups. Note whether memory was recently added, whether modules are matched, and whether firmware uses default settings. On a warranty device, contact the manufacturer before opening it. Qualified technicians can test modules and slots individually using vendor procedures.

If no errors are detected, repeat the original workload and continue investigating drivers, storage, thermals and power. Do not run endless passes while the machine is overheating or storage is failing.

Escalate physical memory work safely

Opening a laptop can damage clips, void service terms or expose a swollen battery. Static discharge can damage components. If you are not trained, use manufacturer service. Desktop memory reseating also requires complete shutdown, power removal and anti-static precautions.

Repeated memory errors, inability to complete the test, crashes during backup or visible physical damage justify professional assessment.

Common mistakes to avoid

• Treating a clean pass as proof memory can never fail.
• Replacing RAM solely because one crash code contains the word memory.
• Running repeated tests while the PC overheats or the drive loses data.
• Opening a laptop without warranty, battery and static-safety considerations.