PC Has Strange Graphic Artifacts

Could be the video card... it has that classic "video memory corruption/GPU overheating" look to it. Could it be a down-sampling issue with your 4K monitor/DisplayPort? I would rarely blame a monitor for such things but I would give another monitor a shot just to see if it is repeatable there - 4K and the processor chips for those are so new there are bound to be silly issues like this regarding timing and pixel clocks and whatnot.
All in all, if you can ramp your GPU up while in an advanced OS environment with drivers without issue, well, I would just chalk it up to a software issue(Win 10 PE) or basic video driver or BIOS issue (Maybe a fix gets released in the future).
 
So it occurs even within the BIOS, and EVGA recommends ....cleaning out old Windows drivers?

If they want to impress, they should send you a new card, and pay for shipping to get yours back....
 
Is that a default clocked EVGA 970 or is that one of those cards that come already overclocked? I'm guessing its the later which is a complete and utter scam. I have seen many many times where these preclocked cards suffer from instability because they are not properly tested. You can try downclocking it to default clockspeeds and see if you run into the problem while its downclocked.

Edit: Oh I forgot to mention if I was going to try to force the issue to happen again I would try furmark. Yes I realize it happened in bios and during windows installation but its still possible its the OC.
 
Last edited:
Op, what power supply are you running that card with btw?

I would lean toward the card however seeing how it happens even in bios. Once set up a system with a GTX 770 for a client, got it all built, installed and started getting graphics weirdness, ended up pulling the card and returning it, no issues after that.
 
With that power supply man I think you have a bum card. Or the slot is flaking out, but doubt that. I'd try to RMA the card.
 
Bought them because they had a rebate, otherwise I would have been a-OK with the base model. Given that the graphical error occurs when booting up, it doesn't make much sense that an overclock would be to blame especially given that I can game at 4K ultra with no issues and the clock working as hard as it can with max clock rates.

More likely than the ram test you did =P Did you try furmark to see if you could trigger the issue in windows?

Edit: does the issue ever happen on boot when you just have one card in there?
 
Slight warm-up anomaly, potentially only visible with that motherboard and/or PSU? (but, without another 970 card to replace it with, or, another system to test said card in, it will be hard to definitively point fingers at a culprit)
 
Can you test for this anomaly with a single GTX970 installed (alternating each, testing each for artifacts-inducing behaviour?)
 
Don't always be so keen ok waiting till next gen. Sometimes yes sure they may be faster, but look at when amd released the r series 200 cards. For example the r9 280 was effectively the same card as the offer radeon 7950. Besides if you get a replacement can't you use warranty on that?


If you have a spare card of any kind, pop that in and when it does not act up, tell them look, all other components test fine, and things only happen when your card is installed. If you want to twist their arm tell them you are a technician not just an average home user, and that you have also consulted with a lot of other techs who ask have reached the same conclusion you have. You like their products and would rather give them a chance to work it out than to share stories if they won't work with you.
 
Do you have a different monitor to see if the same happens there? Or does the card have a different (non-DisplayPort) output you could try? I'm wondering if it's a downsampling or cabling issue - particularly if the monitor isn't DisplayPort input, don't you need active conversion to go from DisplayPort to HDMI at that resolution?

Edit: Never mind on the ports - I didn't read the monitor specs first.
 
Back
Top