Strange colored dots, flickering triangles, blocky textures, or random lines across the screen are called artifacts, and they mean the graphics card is producing bad data. Artifacts can come from an unstable overclock, overheating memory, a driver problem, or, in the worst case, failing hardware. The good news is that several causes are fixable at home, and a logical process quickly narrows down which one you are facing.
Software or hardware?
The first question is whether artifacts appear only in games or everywhere, including the desktop and BIOS. Glitches confined to certain games often point to drivers or a game bug, while artifacts on the desktop and in BIOS point squarely at hardware. Establishing this early saves a lot of wasted effort, because it decides whether you reach for software fixes or start swap-testing parts.
Cause and first action
| Pattern | Likely cause | First step |
|---|---|---|
| Only after overclocking | Unstable GPU/VRAM OC | Reset to stock clocks |
| Worsens as card heats up | Overheating memory | Improve cooling, check pads |
| One game only | Driver or game bug | Clean driver reinstall |
| Desktop and BIOS too | Failing hardware | Test in another system |
Step by step
- Reset any GPU core and memory overclock to stock; unstable VRAM overclocks are a top cause of artifacts.
- Do a clean driver reinstall, fully removing the old driver before installing the latest stable version.
- Monitor GPU core, hotspot, and memory-junction temperatures under load; overheating VRAM can artifact well before the core throttles.
- Verify the game’s files in case corrupted assets are the culprit.
- Reseat the card and its power cables to rule out a poor connection.
- If artifacts show on the desktop and in BIOS, test the card in another PC to confirm hardware failure.
When it is the hardware
If a clean, stock card artifacts everywhere and in another system, the memory or GPU die is likely failing. Sometimes aged thermal pads let VRAM overheat, and a repad fixes it; other times the silicon is simply dying. A card under warranty with these symptoms should go back to the manufacturer rather than be tinkered with.
Do not confuse artifacts with other display faults
Several problems look like GPU artifacts but are not. A single vertical or horizontal colored line that never moves can be a dying panel rather than the GPU. Brief full-screen flicker on a variable-refresh display is often VRR flicker, a panel behavior, not corruption. A damaged or loose display cable can produce sparkles and dropouts that vanish when you swap cables. Before you condemn the graphics card, try a different cable and a different monitor or input. Ruling out the display and cable takes five minutes and can save you from returning a perfectly good GPU, or from ignoring a genuinely failing one.
FAQ
Can a factory-overclocked card artifact at stock speeds?
Yes, if its factory overclock is marginal or the card runs hot. Try lowering the core and memory clocks slightly below stock; if artifacts vanish, the card was on the edge, often because of heat rather than a bad chip.
Do artifacts always mean my GPU is dying?
No. Unstable overclocks, bad drivers, corrupt game files, and overheating VRAM all cause artifacts and are fixable. Only persistent artifacts at stock speeds that appear on the desktop and in BIOS strongly suggest failing hardware.
Bottom line
Diagnose artifacts by asking where they appear. In-game-only glitches usually trace to unstable overclocks, hot VRAM, or drivers, all of which you can address by resetting clocks, improving cooling, and doing a clean driver reinstall. Rule out the cable and monitor first, since those mimic GPU faults. Artifacts on the desktop and in BIOS, especially in a second system, point to failing hardware and a warranty claim. Work from software to hardware, and you avoid replacing a card that just needed a stock reset.
