How To Test If My GPU Is Working Flawlessly?

[jippa_lippa]

Hello.
I recently installed a Gigabyte GTX 980 (Windforce) inside my case.
Problems with the case (even though i have a modern full tower) caused me to accidentally hit the card once, and some "constrictions" in cabling are forcing the power supply pins (the PCI-E ones) a bit.
Paranoia-mode started (considering i spent all my savings on the card).
Other than that, I'd like to know if there's a reliable test i can run to ensure the graphics card works flawlessly and doesn't have any errors (i can even purchase a program, if really worth it).
Simple question as as that.


Small note: One time i used this software: http://www.ocbase.com/index.php/download
Is it reliable?

Thanks a lot
I really appreciate your help

 

[APassingMe]

Use a GPU stress test and [strike]run it for 24 hours [/strike]to see if there are any flickers or crashes. Most people use furmark since it's free, here's a link.

http://www.ozone3d.net/benchmarks/fur/


Edit: It seems that those more experienced than me in GPU testing think that 24 hours is too long for a furmark burn-in and since I normally test CPUs and not GPUs as a rule, I am going to bow out to their experience on this one and won't suggest run time.

 

[WildCard999]

Any current or next-gen game would be fine to test it. GPU-Z is a really good program to make sure the card is not over-heating or any voltage problems and its free. Also just make sure the card isn't bending and nothing is blocking the card from getting air and make sure the pins (wires powering the card) are properly seated. If no issues after that you should be fine. I would also reccomend registering the card in case of issues down the road. Most companies will replace the card within a year on registering.

 

[jippa_lippa]

Any current or next-gen game would be fine to test it. GPU-Z is a really good program to make sure the card is not over-heating or any voltage problems and its free. Also just make sure the card isn't bending and nothing is blocking the card from getting air and make sure the pins (wires powering the card) are properly seated. If no issues after that you should be fine. I would also reccomend registering the card in case of issues down the road. Most companies will replace the card within a year on registering.

The problem is that the 2 "pins sockets"" are a bit offset from each other (the pins enter correctly though).
I'm overly worried...i know...
Anyway is it safe to think the card has no flaws, if it doesn't show any problem after about a week of intensive gaming?

 

[smeezekitty]

24 hours is generally way more than you need. I would try running an intense game or benchmark.
Even one of the unigine things would be fine. Watch your temps and watch for crashes. If it is all ok then don't worry.

 

[Mechanics]

I suggest you do not run these "stress test" for so, 24 hours.

Also just run furmark, or performancetest.

 

[Deleted member 362816]

Use a GPU stress test and run it for 24 hours to see if there are any flickers or crashes. Most people use furmark since it's free, here's a link.

http://www.ozone3d.net/benchmarks/fur/

Running Furmark for 24 hours is like running a car without oil, The risk of damage to the gpu is very high with furmark anything over 30 minutes on the stock cooler. I never ran over 3 hours with a liquid cooled card.

 

[APassingMe]

Use a GPU stress test and run it for 24 hours to see if there are any flickers or crashes. Most people use furmark since it's free, here's a link.

http://www.ozone3d.net/benchmarks/fur/

Running Furmark for 24 hours is like running a car without oil, The risk of damage to the gpu is very high with furmark anything over 30 minutes on the stock cooler. I never ran over 3 hours with a liquid cooled card.

I don't know about the comparison to running a car without oil... the card should throttle before heat becomes a real issue. But yes, it will put a lot of stress on the card and if you're okay with a shorter run time then please go with that as it will be easier on the card overall.

I would probably compare 24 hours on Furmark to a couple of months of heavy gaming so... yeah 3 hours or less is probably enough, but if I really wanted to eliminate the GPU as having any issues this is probably what I would do. Good luck figuring things out.

 

[colmcd]

A lot of these answers seem to be "do a burn in", which implies a GPU burnin.
Naturally your GPU has both a processor and ram. Sadly a fair few Burnin's are basically thrashing the processor but not testing the ram.

Not 100% experienced in GPU testing. But in testing systems I want to test both the Processor, Ram and ensure that were using latest Firmware. (or if I have just upgraded firmware and experiencing problems... Downgrade).

Found this helpful as it points to 4 memory testers as well as load testers.
https://www.raymond.cc/blog/having-problems-with-video-card-stress-test-its-memory/

I suspect my problems on a newish lenovo laptop relate to a dual video card (intel 620/Nvidea mx920). For which I am about to throw this i7 8gb 7500u laptop in the bin!

Seriously hope someone else has more info here.