Log Dejargonizer

Windows System Log · Event 4101

Event 4101: Display driver stopped responding and recovered

Your graphics card stopped answering for a couple of seconds, so Windows reset it. That is the black flicker you just saw. One is nothing. Several a day means the driver, the card, or its power supply needs attention.

Also written as Event ID 4101nvlddmkm stopped respondingTDR timeoutatikmpag stopped responding

What it means for you

On a personal computer

Your screen went black for a second and came back, and maybe a game or video stuttered. Windows noticed the graphics card had stopped responding and restarted it rather than crashing the whole machine. If this happens once in a while, ignore it. If it happens while gaming or watching video most days, work through the driver and cooling first.

For an analyst

This is Timeout Detection and Recovery. The GPU failed to complete work within the timeout — two seconds by default — so the display stack reset it. On virtual desktop hosts it usually means GPU contention rather than hardware. Repeated TDRs frequently precede a 0x116 VIDEO_TDR_ERROR stop error.

What it looks like

Sanitised. Addresses come from the RFC 5737 documentation ranges.

Sample
Display driver nvlddmkm stopped responding and has successfully recovered.

The fields that decide it

Everything else in the log line is context.

Driver name
Which display driver reset, and the strongest clue about the vendor.
  • nvlddmkmNVIDIA. By far the most common name in this event.
  • amdkmdap or atikmpagAMD.
  • igdkmd64 or igfxIntel integrated graphics.
Recovery outcome
Whether the reset succeeded. 'has successfully recovered' means Windows fixed it and nothing was lost beyond a flicker.
How often it repeats
Not a field, but the thing that decides whether it matters. Once a month is noise; several a day is a fault.
What you were doing
Also not a field. TDRs that cluster around gaming, video playback, or GPU-accelerated work point at load, heat, or power rather than a random driver bug.

Ordinary reasons this happens

Most of the time it is one of these.

  • A one-off driver hiccup, which happens on healthy machines occasionally.
  • A driver update installing while the display was in use.
  • A demanding game or application briefly overwhelming the card.
  • Browser hardware acceleration conflicting with an older driver.
  • GPU contention on a virtual desktop host, where the guest sees a timeout it did not cause.
  • Multiple monitors at high refresh rates pushing an older card harder than it likes.

When it is not ordinary

An unstable or malicious kernel driver destabilising the graphics stack.

What gives it awayRepeated TDRs starting shortly after unfamiliar software was installed, particularly alongside other driver load failures.

ATT&CK T1547.006

Cryptomining consuming the GPU.

What gives it awayPersistent TDRs on a machine that is otherwise idle, with sustained high GPU load and an unfamiliar process holding it.

ATT&CK T1496

What to do next

  1. Count how often it happens before doing anything. One a month needs no action at all.
  2. Note what you were doing each time. Clustering around games or video points at load, heat, or power.
  3. Do a clean driver reinstall rather than an in-place update — remove the existing driver fully, then install the current one from the card manufacturer, not from Windows Update.
  4. If it started after a driver update, roll back to the previous version. Newer is not always better for older cards.
  5. Check temperatures under load and clean out dust. Thermal throttling and TDRs go together.
  6. On a desktop, confirm the power supply is adequate and the card's power connectors are properly seated.
  7. Only after all of that, consider raising the TDR delay in the registry — it hides the symptom rather than fixing anything, and it makes real hangs last longer.

Queries to run

powershell How often and how recently. The count is the whole diagnosis. 
Get-WinEvent -FilterHashtable @{LogName='System'; Id=4101} -MaxEvents 50 | Select-Object TimeCreated, Message | Format-List
powershell Daily counts over a month, which makes a worsening trend obvious. 
Get-WinEvent -FilterHashtable @{LogName='System'; Id=4101; StartTime=(Get-Date).AddDays(-30)} | Group-Object { $_.TimeCreated.Date } | Select-Object Count, Name
powershell The installed driver version and date. Compare against the current release from the card manufacturer. 
Get-CimInstance Win32_VideoController | Select-Object Name, DriverVersion, DriverDate, AdapterRAM
wevtutil 
wevtutil qe System /q:"*[System[(EventID=4101)]]" /f:text /c:10 /rd:true

Common questions

What does 'display driver stopped responding and has recovered' actually mean?

Your graphics card did not finish a piece of work within the time Windows allows — two seconds by default. Rather than let the machine hang, Windows reset the display driver and carried on. The black flicker is that reset. Nothing was damaged, and 'successfully recovered' means it worked.

How do I fix Event 4101?

Start with a clean driver reinstall: fully remove the existing display driver, then install the current one from NVIDIA, AMD, or Intel directly rather than through Windows Update. If it began after an update, roll back instead — newer drivers regularly break older cards. Then check temperatures and dust, and on a desktop confirm the power supply and the card's power connectors.

Should I increase the TdrDelay registry value?

Only as a last resort. It gives the card longer before Windows resets it, which hides the symptom without addressing the cause and makes genuine hangs last longer. Work through drivers, cooling, and power first — those fix the problem rather than masking it.

Is Event 4101 a sign my graphics card is dying?

Not usually. Driver problems cause far more of these than failing hardware. Hardware becomes the likely explanation when it happens under load with high temperatures, when you also see visual artefacts, or when it persists across a completely clean driver reinstall and a different driver version.

Why does this happen when I play games or watch video?

Because that is when the card is under real load. TDRs cluster around demanding work because that is when a card that is too hot, underpowered, or running a buggy driver is most likely to miss the timeout. Clustering is a useful clue: it points at load, heat, or power rather than a random fault.

Read next

Vendor documentation

Last reviewed 28 August 2026