You're gaming or working with graphics software and something looks wrong. Strange coloured pixels scattered across the screen. Diagonal lines splitting the image into offset chunks. Flickering textures or entire sections of the display showing corrupted colours. These visual glitches are called artefacts, and they're your GPU telling you something is wrong.
Understanding what type of visual issue you're seeing tells you a lot about the cause. Let's break down the different types and what each one means.
Screen Tearing: What It Is and What Causes It
Screen tearing looks like a horizontal line splitting the image — the top half of the screen shows one frame and the bottom half shows the next frame, slightly offset. This happens when the GPU renders frames at a rate that doesn't synchronise with the monitor's refresh rate. It's not a sign of hardware failure — it's a sync issue.
Fix it:
- Enable V-Sync in your game's graphics settings. V-Sync forces the GPU to synchronise its output with the monitor's refresh rate, eliminating tearing. The trade-off is some added input lag.
- Use G-Sync (NVIDIA) or FreeSync (AMD) if your monitor supports it. These adaptive sync technologies eliminate tearing without the input lag penalty of V-Sync. FreeSync is free and supported on many monitors including mid-range ones.
- Cap your frame rate just below your monitor's refresh rate — this also prevents tearing in many cases.
GPU Artefacts: The Concerning Visual Glitches
Unlike screen tearing, artefacts typically indicate hardware problems. The common types:
Random coloured pixels or dots: Small coloured squares or dots scattered across the screen — sometimes called "snow." This is one of the most classic signs of a failing or overheating GPU. Sustained high temperatures cause the GPU's memory chips or shader units to produce incorrect data.
Corrupted textures: Surfaces in a game appear as solid wrong colours, patterns, or stretched textures. This indicates GPU VRAM (video memory) issues — either overheating, hardware failure, or insufficient VRAM for the resolution/texture quality you're running.
Triangles or polygons stretching across the screen: Geometry suddenly stretches from its normal position to the edge of the screen. This is a GPU shader or VRAM failure symptom, especially under heavy load.
Flickering display or black flashes: The screen goes black briefly, then returns to normal. This can indicate a failing display connection (cable, port), a driver issue, or a GPU struggling to maintain stable output.
Diagnosing GPU Artefacts
Step 1: Check Temperature
Install MSI Afterburner and run a GPU-intensive task. Monitor GPU temperature and GPU memory temperature. A GPU running above 85°C or GPU memory above 90°C under load is thermal throttling and potentially suffering thermal damage. If temperatures are high, clean the GPU's heatsink, ensure case airflow is adequate, and consider replacing the GPU's thermal pads (a more advanced procedure).
Step 2: Check and Update (or Roll Back) Drivers
GPU driver bugs can cause artefacts that look exactly like hardware failure. Before concluding the GPU is dead, install the latest driver from NVIDIA or AMD's website, or roll back to a previous stable version using Display Driver Uninstaller (DDU) to perform a clean installation.
Step 3: Test at Lower Settings
If artefacts disappear when you reduce texture quality or resolution, the GPU's VRAM is being overwhelmed (running out of memory). This isn't hardware failure — it's a settings issue. Either reduce texture settings or upgrade to a GPU with more VRAM.
Step 4: Check the Display Connection
Flickering and some artefacts can come from a faulty cable rather than the GPU itself. Try a different HDMI or DisplayPort cable. Try a different port on the GPU. Try a different monitor if possible.
Step 5: Test the GPU in Another System
If artefacts persist across different drivers, cables, and monitors, and temperatures are acceptable, the GPU hardware itself may have failed. Testing in another PC confirms this. If artefacts appear in the other PC too, the GPU needs replacement.
Nigerian Context: Power Damage and Overclocking
In Nigeria, two factors cause GPU artefacts more frequently than elsewhere:
- Power surge damage: Voltage spikes can partially damage GPU silicon or VRAM chips, causing artefacts that appear progressively. A GPU that was working fine before a power event and shows artefacts after is almost certainly surge-damaged.
- Tokunbo GPUs: Second-hand GPUs sold in Computer Village and similar markets often come from unknown sources, sometimes from crypto-mining rigs that were run hot 24/7 for years. A tokunbo GPU showing artefacts was likely run into early degradation by its previous owner. Always test a used GPU thoroughly before buying.
When to Replace the GPU
If artefacts persist after updating drivers, temperatures are within normal range, and the GPU shows artefacts in another system, replacement is the right call. Using a GPU with hardware failure long-term risks sudden complete failure and, in some cases, damage to the PCIe slot on the motherboard.
Looking for a GPU upgrade? Build your next system with us →
Not sure if your GPU needs replacing? Our team can test it →