Take a photo or screenshot of the artifacts and note when they appear (always, only in one game, only when hot). Back up important files if the PC also crashes. If Windows cannot start, see the startup guide.
Work through these checks
0 of 7 complete01Find out where the artifacts appear
This one check splits the problem in two: software and settings, or hardware.
- Restart the PC and watch the manufacturer logo and the BIOS setup screen (usually Del or F2).
- Check the Windows sign-in screen and the desktop without any game running.
- Then check whether the artifacts appear only in one game, in all games, or after the PC has been under load for a while.
Artifacts only in games or after heating up: start with steps 3 to 5. Artifacts already in the BIOS or at startup: the card is likely faulty; still do steps 2 and 6, then step 7.
If it does not helpIf only one game shows artifacts, update or repair that game (most launchers have a Verify files option) before changing anything else.
02Rule out the cable and monitor
A damaged cable or a failing monitor can produce sparkles, lines or flicker that look exactly like GPU artifacts.
- Try a different DisplayPort or HDMI cable and a different port on the graphics card.
- If possible, connect another monitor or a TV.
- Take a screenshot (Win + Shift + S) while artifacts are visible and look at it on another device.
If the screenshot is clean but the screen is not, the problem is after the GPU: cable, port or monitor. If the screenshot also shows artifacts, the GPU is drawing them.
If it does not helpReplace the cable or have the monitor checked if the GPU is not the cause.
03Remove overclocks and tuning profiles
Video memory overclocks often cause artifacts before they cause crashes. Factory OC modes in vendor software count too.
- In MSI Afterburner, AMD Software, NVIDIA App or the card maker's utility, reset clocks, voltage and power limit to default and disable profiles that load at startup.
- Turn off any undervolt or custom fan curve while testing.
- Restart and test the game or app that showed the artifacts.
Artifacts caused by tuning disappear at default settings.
If it does not helpIf you want the tuning back, increase it gradually and stop well below the point where artifacts appeared.
04Clean-install the graphics driver
Corrupted driver files or leftovers from an older driver can cause rendering errors.
- Download the latest driver for your exact card from NVIDIA, AMD or Intel, or from your laptop maker.
- Run the installer and choose the clean installation or factory reset option if it offers one.
- If the artifacts started right after a driver update, roll back instead: Device Manager > Display adapters > right-click the card > Properties > Driver > Roll Back Driver.
- If Windows is still running but the picture freezes or goes black, press Win + Ctrl + Shift + B to restart the graphics driver.
Download drivers only from the GPU or PC manufacturer.
Driver-related artifacts are gone after the reinstall or rollback.
If it does not helpIf the screen is unusable, start Windows in Safe Mode (it uses a basic display driver) and reinstall the driver from there.
05Check GPU and memory temperatures
Video memory overheats more easily than the GPU core, and hot memory produces artifacts that appear after some time under load.
- Run a monitoring tool such as HWiNFO or GPU-Z while gaming and note GPU temperature, hot spot and, if shown, memory junction temperature.
- Compare them with the limits for your card model and check that all GPU fans spin.
- Clean dust from the card's heatsink and improve case airflow.
Opening the graphics card's cooler can void its warranty.
Temperatures stay below the card's limits and the artifacts do not return when it is cooler.
If it does not helpIf temperatures reach the limit even when the card is clean, the thermal pads or paste may need replacing; for a card under warranty, ask the manufacturer before opening it.
06Check power connectors and the power supply
A card that does not get stable power can produce artifacts and then crash under load.
- Shut down and unplug the PC.
- Make sure every PCIe power connector on the card is pushed fully in. A 12VHPWR/12V-2x6 connector must sit completely flush with no gap.
- Use separate cables from the power supply for each 8-pin connector where the card's manual recommends it, and check that the power supply meets the card's requirements.
- Reseat the card in its slot and check that it is screwed to the case.
The card is firmly connected and properly powered.
If it does not helpIf you see melted or discolored plastic on a power connector, stop using the card and contact the manufacturer.
07Test in another PC or claim warranty
Artifacts at default settings, with a clean driver, normal temperatures and good power, mean the card itself is failing, most often its video memory.
- If you can, test the card in another PC, or a known-good card in yours.
- Check the warranty with the card maker using its serial number, and include your photos and test results in the claim.
You know whether the card or the rest of the PC is at fault.
If it does not helpUntil it is replaced, use the processor's integrated graphics if it has them, by connecting the monitor to the motherboard's video output.
Possible causes
- GPU or video memory overclock, including factory OC profiles set in software
- Video memory (VRAM) or GPU overheating
- Corrupted or incompatible graphics driver
- Loose or underpowered PCIe power connectors
- Faulty cable or monitor (looks like artifacts, but is not the GPU)
- Failing graphics card, usually its video memory
Useful Windows tools
These commands open Windows tools. Review changes before applying them.
dxdiagdevmgmt.mscmsinfo32Was this helpful?
Your progress is saved on this device.