Not sure how to tell if your graphics card is on the verge of death or not? Thankfully, it’s one of the easier components to diagnose for any errors. There are plenty of warnings signs, too! In this article, we’re going to show you how to look out for those warning signs, troubleshoot problems and hopefully get to the bottom of what’s happening.

Be sure to follow along below.

Editor’s note: This guide was originally posted in September of 2014, but we have brought it up to date as of November 2016, including a whole lot more ways to troubleshoot video card problems.


Since the graphics card has to deal with almost everything “visual” you see on on your computer, warnings that the card is on its way out the door can be really obvious. Below are some early warning signs of video card failure.

  1. Stuttering: When a graphics card is going bad, you can see quite a bit of stuttering beforehand. You shouldn’t use this to determine whether your graphics card is bad or not, though. Malware, a dying hard drive and even RAM problems can all cause this same issue; however, if you start seeing other warning signs to go with this one, there’s a good chance it’s your graphics card.
  2. Screen glitches: Other times, you’ll see the screen glitch out. If you’re playing a game or watching and movie and suddenly start seeing tearing or weird colors appearing all over the screen, your graphics card might be dying. Sometimes if you restart your computer, the screen will go back to normal, but expect the same problem to come back if you have a faulty graphics card.
  3. Strange artifacts: Similar to screen glitches, a bad graphics card can usually result in strange artifacts all over your screen (example here). This can sometimes be fixed by a restart, but once again, if you have a faulty graphics card, expect the problem to come back. It’s worth noting that the cause of this usually comes from excessive overclocking (learn about proper overclocking here), heat problems and even too much exposure to dust.
  4. Blue screens: Blue screens is something we’re all familiar with. A computer can blue screen for a number of reasons, whether that be problems with RAM, hard drives, graphics cards or other components. But, if the system crashes and/or blue screens when you start doing some graphic intensive tasks (e.g. video games, watching movies, etc), this could be an indication your graphics card is on its way out.
  5. Fan noise: This does not necessarily correlate to needing to replace your graphics card, but keeping an eye out for a louder-than-normal fan noise on your graphics card can indicate the card is getting too hot. If it’s getting too hot, you’ll want to stop what you’re doing and try and clean it out as best as possible. If the fan isn’t able to quiet down, it’s possible that something is internally wrong.


sapphite-ati-radeonAs we always mention in our troubleshooting guides, finding out what’s wrong and diagnosing a problem is always a process of elimination. That said, starting with checking your connections is the best place to start. Loose connections can cause a lot of problems, especially with a graphics card.

In some cases, you won’t be able to check connections, particularly if you have a laptop, especially from specific manufacturers like Alienware who make it a bit more difficult to access components.. That’s, of course, not a problem, as — generally speaking — you won’t have an issue with loose connections in a laptop. With laptops, more often than not, the problem is dust because of being in such an enclosed space. If you can open it up and clean out any dust you can, that would be the first place to start. If dust or lint has been in there for an extended period of time, it can easily fry a component or cause it to overheat by not giving it enough air or letting the fan run properly.

heaven-benchmark-pageThe next thing you can do is run some software tests. As far as diagnostic tools go, there’s not a whole out there that’s reputable or even reliable. Your best bet is to run something like GPU-Z and watch the real-time temperature for any oddities. For actually testing the card, there’s nothing like putting it through some real-world use. You can run a video game or a really visually intense movie. Alternatively, you can use the Heaven Benchmark tool to test your card. Run it for a couple hours — it should be able to handle it without crashing or causing any graphical errors like strange artifacts and stuttering.

It’s also worth noting that if you don’t have a graphics card and are using a motherboard’s integrated graphics, this could be a sign of motherboard failure on the way instead. Be sure to check out our troubleshooting guide for motherboard failure.

Next, make sure your drivers on your graphics card (and monitor) are all up to date. You can also try uninstalling the ones you already have and then re-install them to ensure there aren’t any problems there. It’s worth noting that you can uninstall your drivers without losing video. Once uninstalled, Windows will use some very basic drivers to display video to your monitor. So, you won’t actually lose video functionality or cause any harm to the card. But, as always, be sure to consult your video card’s manufacturer for specific uninstall/reinstall instructions. You can find some specific instructions from NVIDIA and AMD here and here, respectively. AMD actually has a free cleaning tool to do this for you automatically.

If you were to take your computer to a repair shop, they would simply swap out the graphics card for another graphics card to see if that’s where the problem lies. If the test graphics card they put in allows the computer to operate without an issue, it’s obvious the old graphics card needs to be replaced. If you don’t mind fiddling with components at home and have an extra or cheap graphics card lying around that will fit your computer, you can do the same test process yourself. You’ll simply need to swap out the graphics card and see if the problem persists with the new one you just put in. If so, it’s time for a replacement card.

It’s also worth checking for any physical problems. If the fan has stopped working on the video card or you see any leaking or bulging capacitors, it’s time for a replacement. In cases of this happening, usually the video card will stop working almost immediately.

In some cases, the problem could lie with a virus or piece of malware on your computer. That’s likely not what’s causing strange artifacts or screen glitches, but if you’re getting some stuttering or experiencing frequent crashes, there’s a good chance malware is the culprit. Be sure to run your anti-virus software, and to be extra sure it’s not something in the system files, you should run some bootable anti-virus software (Bitdefender has an excellent tool for just that).

What causes video card failure?

dusty-gpu-fanVideo cards can fail for so many different reasons. Not properly installing the component in the computer can lead to video card failure, but more commonly, dust and lint are the culprit. Dust itself generally isn’t the problem, it’s more that they block fan vents and prevent proper cooling. In some cases, if bad enough, dust can actually insulate a component and cause overheating that way.

Some other things that can cause video card failure is too much overclocking, as mentioned above. Overclocking at the stock voltage is more than safe. If you push the card to its limits with high voltage, that will kill a card sooner than normal. But, even that will take months or years to kill a card. It’s also worth noting that many modern cards a pretty resilient to excessive heat, but do keep in mind that this can put extra wear and tear on the card, and even eventually fry it if the heat output is greater than what your heat sink can handle.

Aside from that, the last thing that can kill your video card is the standard electrical outage. Blackouts, brown outs and power surges can fry all of the components in your computer — even the graphics card. In most cases, if you have some extra cash to spare, you can prevent this situation. All you’ll need to do is invest in a quality surge protector as well as a Uninterruptible Power Supply (UPS). As you know, the primary role of a UPS is to provide temporary power in case the source is cut off so that you can properly shut down your machine; however, it’s also able to help prevent damage from things like power surges. You can read more about what a UPS and surge protector does here.

Ultimately, the video card is subject to as much wear and tear as anything else. If your card fails, it may have just been time for the card to fail. In that case, a replacement is your only choice.

Replacing your video card

Now, if you’re finding yourself needing a replacement, we have quite a few options. Depending on the type of work you’re doing, you don’t necessarily need a super expensive video card. If you’re on a budget, we’ve got a great guide on buying a graphics card for almost any price range. But, before going out and buying a new card, there are a few things to look at and find out what you need, such as clock speed and memory size. PCMech’s very own Christian de Looper put together a great guide on what types of things you should look out for for your own build.


Video cards are generally pretty easy components to diagnose. If you can, it’s as simple as swapping out the graphics card for another and seeing if the problems remain. But, there are a few extra steps you can take to fully check things out, as we covered above. If a video card completely fails, you’ll notice immediately, as you just won’t have any video. That said, if you do experience video card failure, hopefully you have your motherboard’s onboard graphics to rely on until you can pick up a replacement component.

We hope this guide helped you get to the bottom of your problem. But if you’re still stuck, be sure to head on over to the PCMech forum and post your problem to get some additional help from the PCMech community!