Graphics Card Not Working A Complete GPU Troubleshooting Guide

Quick Answer

When a graphics card stops producing a display or crashes under load, the card itself is only one possible cause. A monitor, cable, driver, power supply, motherboard setting, unstable overclock or cooling problem can create similar symptoms.

The safest approach is to change one variable at a time and record what happens. This guide starts with quick external checks, moves through Windows and driver diagnosis, and only then covers internal hardware testing.

Graphics Card Not Working? Start With the Symptom

Match the symptom to the stage at which it occurs before opening the PC or removing drivers. This prevents a display problem from being treated as a hardware failure and stops software troubleshooting from wasting time when the card is not initialising.

A useful distinction is whether the failure happens before or after Windows takes control. If the computer never shows its startup logo, start with the display path, power and hardware. If the image disappears only after Windows loads, software and driver checks move higher up the list.

Before You Begin: Protect the PC and Yourself

Start with non-invasive checks. Only open a desktop PC if you are comfortable working around internal components and the device is no longer connected to mains power.

Follow these precautions:

1. Shut down the PC, switch off the power supply and unplug it before touching the graphics card or internal cables.
2. Work on a stable surface and discharge static by touching an unpainted metal part of the case before handling components.
3. Never open a power supply. Components inside can retain dangerous electrical charge after the PC is unplugged.
4. Never reuse a modular PSU cable from another power supply unless the PSU manufacturer explicitly confirms compatibility. Identical-looking plugs can be wired differently.
5. Do not flash a motherboard BIOS or graphics card firmware as a general troubleshooting experiment. Use firmware only when the manufacturer documents a relevant fix for the exact model.
6. Avoid removing a cooler, thermal pads or thermal paste while the card is under warranty unless the manufacturer has approved the work.

Step 1: Record When the Graphics Card Stopped Working

Write down what changed immediately before the problem began. Useful details include a Windows update, new graphics driver, new monitor, GPU installation, PSU replacement, BIOS reset, overclock, case move or internal cleaning.

Also note the exact failure point:

• Before the startup logo
• While Windows is loading
• After signing in
• When a game starts
• After several minutes under load
• When waking from sleep
• Only when using a particular monitor, port or application

This timing can save substantial work. For example, a display that works in the BIOS but turns black when Windows loads is more consistent with a driver or display-mode problem than a completely unpowered card.

Create a short troubleshooting log as you work. Record the change, the result and whether the symptom occurred at the same point, because repeatable evidence is more useful than a single successful startup.

Step 2: Rule Out the Monitor, Cable and Display Path

A GPU black screen does not always originate from the GPU. The monitor may be on the wrong input, the cable may have failed, or Windows may be sending the image to a different display.

Check the following in order:

1. Confirm that the monitor has power and is set to the correct HDMI, DisplayPort or USB-C input.
2. Disconnect extra displays and test one monitor at a time.
3. Reconnect the display cable at both ends.
4. Try another cable that is known to work.
5. Test a different output on the graphics card.
6. Connect the PC to another monitor or television if one is available.

On a desktop with a dedicated graphics card, the display cable normally connects to the ports on the card, not the motherboard. A motherboard output uses integrated graphics where supported and can make a healthy discrete GPU appear inactive.

The motherboard output is still useful as a controlled test if the processor has integrated graphics. If integrated graphics produces a stable image while the dedicated card does not, continue with power, detection and driver checks rather than immediately declaring the GPU dead.

If Windows is running but the wrong display mode is selected, press Windows + P, wait briefly, then choose the appropriate screen mode. This will not help when the PC fails before Windows starts.

Step 3: Try the Safe Windows Display Recovery Checks

If the screen froze or went black while Windows was already running, press Windows + Ctrl + Shift + B. Windows should attempt to restart the graphics driver, and the display may flicker or make a brief sound.

However, this shortcut is useful for a temporary graphics-stack freeze. It cannot repair loose hardware, inadequate power, excessive heat or a damaged card.

For a black screen with a visible pointer, try Ctrl + Alt + Delete. If the security screen appears, open Task Manager and restart Windows Explorer. A desktop-process failure can look like a GPU fault even when the graphics hardware is still producing an image.

Safe Mode is the next useful separation when the display fails as Windows loads. If the picture works with Microsoft's basic display driver in Safe Mode, the installed graphics driver, display configuration or startup software becomes more likely than a total hardware failure.

Step 4: Check Whether Windows Detects the GPU

Open Device Manager by right-clicking Start and selecting Device Manager, then expand Display adapters. The exact entry provides more information than a general statement that the GPU is missing.

Select Action > Scan for hardware changes after a driver change or hardware reconnection. If the device appears only temporarily, note that result rather than repeatedly reinstalling the same driver.

Step 5: Update, Roll Back or Reinstall the Graphics Driver

GPU driver issues often begin after an update, an interrupted installation or a switch between graphics card brands. The correct action depends on what changed.

If the Problem Began After a Driver Update

Try rolling back to the previous driver through Device Manager if the option is available. A newer release can fix one game while introducing a problem with another system configuration, so “latest” does not always mean “most stable for this PC”.

If the Driver Is Missing or Outdated

Download the correct driver for the exact GPU model and Windows version from NVIDIA, AMD, Intel or the computer manufacturer. Avoid third-party driver download sites and do not assume that drivers for similarly named cards are interchangeable.

If a Normal Update Does Not Fix the Problem

Use the manufacturer's standard uninstall or clean-install option before turning to a third-party removal utility. AMD provides its own Cleanup Utility, while NVIDIA users can reinstall the correct Game Ready or Studio driver through the current NVIDIA software and official driver package.

After the installation, restart the PC and repeat the same controlled test. Do not change BIOS settings or overclocking profiles at the same time, because you need to know whether the driver action changed the result.

Step 6: Return Performance Settings to Stock

An unstable GPU overclock, undervolt, memory tune, custom fan curve or power-limit change can produce crashes that resemble hardware failure. Restore the graphics card to its default settings before testing further.

Temporarily return CPU and RAM tuning to standard settings as well. A GPU-heavy game can expose system-memory or CPU instability, so a display-driver crash does not automatically prove that the graphics card caused it.

Disable overlays, recording tools and hardware-monitoring utilities for one test if the issue occurs only in games. If only one application fails, verify its files, update it, rebuild its shader cache where supported and test a different graphics API or lower VRAM-heavy settings before changing hardware.

When several unrelated games and creative applications fail in the same way at stock settings, the evidence points more strongly towards a shared driver, heat, power or hardware problem.

Step 7: Check Desktop GPU Power and PCIe Seating

Skip this section if you have a laptop or all-in-one PC because their graphics processors are normally integrated into the mainboard. Desktop users should only remove a card if they are comfortable handling PC components.

With the PC switched off, unplugged and discharged:

1. Check that the card sits level and fully inserted in the primary PCIe slot.
2. Confirm that the slot-retention latch is engaged and the case bracket is secured.
3. Inspect each GPU power connector for a partial connection, damaged pin, discolouration or melted plastic.
4. Reconnect the required power plugs until they are fully seated, following the graphics card and PSU manuals.
5. Check that a heavy card is not pulling away from the slot because of excessive sag.

Do not sharply bend a high-power cable immediately beside the connector. Also avoid forcing a plug that does not align cleanly, because a mismatched CPU, peripheral or modular PSU cable can damage hardware.

An inadequate or failing power supply may allow the PC to work on the desktop but restart, freeze or lose video during gaming. Compare the complete system against the graphics card manufacturer's PSU recommendation, including the processor and other components, rather than looking only at the wattage printed on the unit.

A Kernel-Power event in Windows confirms that the system shut down unexpectedly, but it does not identify the PSU as the cause. Heat, unstable memory, motherboard faults and interrupted mains power can produce the same event.

Step 8: Review BIOS, UEFI and Laptop GPU Modes

BIOS and UEFI changes come after the simpler checks. Menu names vary by motherboard, so use the manual for the exact computer or board rather than following a generic screenshot.

Relevant settings may include:

• Primary Display, Initial Display Output or PEG/PCIe Graphics
• Integrated graphics or hybrid graphics modes
• Laptop MUX, Eco, Standard, Optimised or discrete-GPU modes
• PCIe slot configuration after a BIOS reset

A laptop power-saving or Eco mode may deliberately hide the dedicated GPU from Device Manager. Change the mode through the manufacturer's control application, connect mains power and restart before treating the missing entry as a fault.

Restoring BIOS defaults can help after an incorrect setting, but record any storage, boot or security configuration first. Update BIOS or GPU firmware only from the official manufacturer support page and only when the release notes address compatibility or the specific problem you have.

Never interrupt a firmware update. If the system is unstable, has unreliable power or cannot maintain a display, professional support may be safer than attempting an update.

Step 9: Check for Graphics Card Overheating and Fan Problems

Graphics card overheating can cause clock-speed reductions, stuttering, driver resets, black screens and shutdowns. However, there is no single temperature that diagnoses every GPU, because limits differ by model, sensor, cooler, workload and ambient conditions.

Monitor the following while reproducing the problem with a normal game or application:

• GPU core temperature
• Hotspot or junction temperature, where available
• Fan speed
• Clock speed
• Power use
• GPU and video-memory use
• Whether performance drops as temperature rises

Stopped fans at idle may be normal because many cards use a zero-RPM mode. A graphics card fan not spinning under sustained load is more concerning if temperature keeps rising, the card throttles or the system crashes.

Check for dust, blocked case intakes, cables touching fan blades and poor exhaust clearance. Hold fan blades still when using suitable compressed air, and do not insert fingers or tools into a powered fan.

If CPU, GPU and case temperatures all rise together, take help from our broader gaming PC overheating guide. That article can cover case airflow, ambient temperature, fan layout and system-wide cooling.

Stop the test if temperature climbs rapidly towards the card's documented limit, the fan fails to respond, artefacts worsen or the PC shuts down. Replacing thermal pads or paste is not a first-line fix and can affect warranty cover.

Step 10: Run a Controlled Stability Test

Do not begin with an extreme stress test. First confirm that the PC is stable on the desktop, during video playback and in a light application, then try a known game or workload that previously worked.

Record the time to failure, temperature, fan response and visible symptom. A benchmark result by itself cannot prove whether the fault sits in the GPU, PSU, RAM, motherboard or driver.

Use this evidence to narrow the cause:

Cross-testing is one of the clearest diagnostic methods, but it requires compatible hardware and a suitable power supply. If you do not have a known-good system or card, a repair technician can perform the comparison without risking another component.

A passed test does not guarantee that every part of a GPU is healthy. It only shows that the system completed that workload under those conditions.

Signs of GPU Failure Vs a Fixable Problem

The strongest signs of GPU failure are repeatable and follow the card after other causes are removed. A single symptom should be treated as a clue, not a verdict.

Visual artefacts can come from a cable, display or unstable settings as well as faulty video memory. Code 43 can also indicate either a driver problem or a device problem, so neither should be treated as automatic proof of a dead GPU.

Laptop GPU Troubleshooting Is Different Than Desktop

Most laptop GPUs cannot be removed, reseated or replaced independently. You need to focus on software, power modes, manufacturer diagnostics, external-display testing and cooling before arranging service.

Useful laptop checks include:

1. Connect the original charger and disable battery-saving modes for the test.
2. Check the manufacturer's hybrid, MUX, Eco or discrete-GPU setting.
3. Confirm which GPU Windows assigns to the affected application.
4. Test the internal screen and an external display separately.
5. Use the laptop manufacturer's BIOS and driver packages where it recommends them for that model.
6. Check whether the fault appears only on battery power or only after the laptop becomes hot.

Do not follow desktop instructions that involve opening the case, changing PCIe slots or replacing PSU cables. A soldered laptop GPU or mainboard fault requires specialist equipment and may be covered by the device warranty.

Common GPU Troubleshooting Mistakes

❌ Changing Several Things at Once

If you replace the cable, reinstall the driver and reset BIOS settings together, you cannot tell which action mattered. Make one controlled change and repeat the same test.

❌ Assuming the Newest Driver Is Always the Best Fix

A current driver may contain necessary support or fixes, but it can also conflict with a particular game or system. If the fault began after an update, rollback is a valid diagnostic step.

❌ Treating Spinning Fans as Proof the Card Works

Fans and lighting can receive power even when the GPU cannot initialise or output a picture. Check Device Manager, the display path and behaviour under load.

❌ Using a Universal Temperature Rule

Core, hotspot and memory limits vary. Compare readings with the specifications for the exact card and focus on throttling, rising temperature, failed fan response and repeatable crashes.

❌ Running Repeated Extreme Stress Tests

A failed stress test does not identify the faulty component by itself. Repeating it on an overheating or unstable system can worsen the problem without producing better evidence.

❌ Flashing Firmware Too Early

BIOS and VBIOS updates are not routine fixes for every black screen. Use them only when the manufacturer identifies a relevant compatibility problem and provides instructions for the exact model.

❌ Ignoring the Rest of the System

RAM instability, a failing PSU, poor CPU cooling, motherboard faults, a damaged display cable and game-specific bugs can all look like a graphics card problem. A good diagnosis tries to disprove these alternatives.

Should You Repair, Claim Under Warranty or Replace the GPU?

Check the warranty before opening the graphics card or paying for a repair. Record the serial number, purchase information, symptoms, driver version, temperature evidence and cross-test results, because the retailer or manufacturer may request them.

A repair can make sense for a replaceable fan, damaged display port or another clearly identified serviceable component. Board-level GPU and video-memory faults require a specialist with appropriate diagnostic and soldering equipment.

Replacement becomes more reasonable when the card fails in another compatible system, another GPU works in the original PC, the warranty has expired and the repair is not economical. Compare graphics cards for gaming on Laptop Outlet by physical size, PSU requirement, display outputs, workload and monitor resolution rather than choosing on model name alone.

The next decision is the GPU platform. NVIDIA GeForce graphics cards may provide better continuity when games or professional applications rely on NVIDIA-specific technologies. AMD Radeon graphics cards may suit different performance, display and feature requirements. Check the exact applications and games you use instead of deciding by brand alone.

Finally, you need to inspect the individual brands rather than relying on the GPU name. MSI graphics cards and ASUS graphics cards based on the same GPU can have different dimensions, cooling systems, noise characteristics and power connectors. These details are particularly important for compact cases and PCs with restricted airflow.

How to Reduce Future Graphics Card Problems

1. Keep a note of the last stable driver instead of updating immediately before important work or a competitive gaming session.
2. Install graphics software from the GPU or computer manufacturer, not a third-party download site.
3. Clean case filters and check airflow before dust blocks the GPU heatsink.
4. Inspect high-power connectors after moving the PC, but only while the system is switched off and unplugged.
5. Support a heavy graphics card if its weight is pulling it away from the PCIe slot.
6. Return clocks and power settings to stock before diagnosing new instability.
7. Avoid placing the PC in a closed cabinet or directly beside a heat source.
8. Record temperatures and symptoms before and after maintenance so improvements can be measured.

Preventive maintenance cannot stop every hardware failure. It can, however, reduce heat, connection and software problems that often make a healthy graphics card appear faulty.

What to Do If Your Graphics Card Is Still Not Working

If your graphics card is not working correctly, the most reliable approach is to identify when the fault begins and eliminate simpler causes before replacing hardware. Check the monitor, cable and output first, then separate Windows and driver faults from power, PCIe and cooling problems. Test one change at a time and use another display, application or compatible system where possible. A single crash, Code 43 message or stopped idle fan does not prove failure. Persistent artefacts, repeated faults at stock settings or problems that follow the card into another suitable PC provide stronger evidence that repair, warranty support or replacement is needed.

What People Also Ask…

Why Is My Graphics Card Not Working Even Though the Fans Spin?

Spinning fans only confirm that the card receives some power. The GPU may still have an incomplete power connection, faulty display path, driver problem, BIOS conflict or hardware fault. Check the monitor, cables, Device Manager and power connectors.

How Do I Know If My GPU Is Dead?

A GPU is more likely to have failed when artefacts appear before Windows loads, it remains undetected in another compatible PC, or several applications crash at stock settings. One black screen, driver error or stopped idle fan is insufficient evidence.

Why Is My GPU Not Detected in Device Manager?

Windows may not detect a GPU because the card lacks power, is poorly seated, is hidden by a laptop graphics mode, or has a driver or BIOS problem. Check Device Manager, power connections and manufacturer-specific settings.

Can I Reset My Graphics Card Without Restarting the PC?

Use the Windows graphics-driver shortcut, Windows + Ctrl + Shift + B, without restarting the PC. The screen may flicker or beep briefly. This can recover a frozen display, but it cannot fix loose hardware, overheating or physical damage.

Can a Graphics Card Be Repaired?

Some graphics cards can be repaired if the problem is a replaceable fan, damaged display port or another serviceable component. Board-level GPU or memory faults require specialist equipment. Compare warranty coverage, repair cost, card age and replacement suitability first.