Little blue pixels/areas when stress testing GPU

[Dunkan77]

Hello,
Are little pixels or areas of the screen a sign of artifacting? I have overvolted the core on my EVGA GTX 970 SC ACX 2.0 by 12 mV to get a more stable OC and these blue patches appeared. Does this mean that the core needs more voltage to run at these frequencies or did I OC too much? Also, does artifacting affect/damage the monitor? My memory is well in it's limits with a 250 MHz offset (with a lower core offset I can set this higher, I've tried) and core clock is 150 MHz offset. I didn't go any further in OC'ing and I'm asking you guys what this is and why is it occuruing.
Thanks for a quick answer so I can resume OC 😀

 

[tambeshakunt]

It can either be the GPU or the monitor. Try using another display to confirm what exactly is producing this issue.

If it is the GPU, what PSU are you using?

And artifacting doesn't usually damage the monitor.

 

[amd4lif3]

i suggest setting clocks volts stock and this will show you if you have a damaged screen or gpu is not stable enough

 

[Khaleal]

It seems you are having artifacts. Over-volting the GPU generates heat on the VRM.. something that causes artifices.
Every GPU has a limit on the voltage.. I can understand that you need to increase to voltage in order to reach higher clocks but when it's overvolted beyond the safe limits it will become unstable..
I suggest lowering the clocks and voltage.

 

[Dunkan77]

It seems you are having artifacts. Over-volting the GPU generates heat on the VRM.. something that causes artifices.
Every GPU has a limit on the voltage.. I can understand that you need to increase to voltage in order to reach higher clocks but when it's overvolted beyond the safe limits it will become unstable..
I suggest lowering the clocks and voltage.

If EVGA set the max OV at 37 mV on the card's BIOS it means that staying within these values is safe right? Also, I heard backplates lower VRM temperatures and should I get one to OC further?

 

[Dunkan77]

i suggest setting clocks volts stock and this will show you if you have a damaged screen or gpu is not stable enough

I checked, it's the GPU that's not stable enough, will this be sorted with more voltage to the core or did I reach the limit here? I have seen people claiming stable 200+ MHz offsets on the core so that's why I'm not happy with 100 MHz less than them, did they set the core voltage offset to max?

 

[Khaleal]

It seems you are having artifacts. Over-volting the GPU generates heat on the VRM.. something that causes artifices.
Every GPU has a limit on the voltage.. I can understand that you need to increase to voltage in order to reach higher clocks but when it's overvolted beyond the safe limits it will become unstable..
I suggest lowering the clocks and voltage.

If EVGA set the max OV at 37 mV on the card's BIOS it means that staying within these values is safe right? Also, I heard backplates lower VRM temperatures and should I get one to OC further?

There is a big difference between a "stable" voltage (the card will run stable without any issues) and "safe" voltage (the card may or may not run stable but it won't cause hardware damage)..
Over-volting generally shortens the life of your card and causes damage to the hardware of your card on the long run.
I have the ASUS R9 280X which is the newer version of the HD 7970GHZ.. Many R9 280X owners complain about random artifacts issues out of the box.. Although mine doesn't do that (well.. it did it one time for some seconds).. This GPU is problematic.. If your GPU is working fine I'd not try to push it further.. It can run all games on ultra with pretty decent fps rates anyway.

 

[amd4lif3]

excuse me was it not understood that you are killing the card just having artifacts im sure lowers the lifespam of your card how can you be not happy (its not dead yet) when its dead then you wont be happy set it back to stock you clearly are not an experianced overclocker overclocking isnt just oh ill copy that person and ill get the same results THAT IS WRONG each graphics card is different and you have to do according to your card. im telling you if you keep this up your card will die you dont want that especcially with a gtx 970 i would cry all day , ya know what im using a gtx 560 pre oc and im sure i can get more out of it but is there a need NO so do i ask for more NO get it into your head overclocking kills JUST LIKE SMOKING gosh you need to set the stock defaults and if you want to oc then watch many tutorials (the detailed ones which explain why you do one thing to have an effect on another thing , i mean im sure you dont need to oc (its a FUCKING GTX 970) secondly you dont need to im sure you will regret damage to your card. im sorry for being so hard on you but so many people nowerdays break there cards because they follow a crummy youtube video which does not explain anything to be good at computers you must MUST know your computer and the limit of each component

 

[tambeshakunt]

excuse me was it not understood that you are killing the card just having artifacts im sure lowers the lifespam of your card how can you be not happy (its not dead yet) when its dead then you wont be happy set it back to stock you clearly are not an experianced overclocker overclocking isnt just oh ill copy that person and ill get the same results THAT IS WRONG each graphics card is different and you have to do according to your card. im telling you if you keep this up your card will die you dont want that especcially with a gtx 970 i would cry all day , ya know what im using a gtx 560 pre oc and im sure i can get more out of it but is there a need NO so do i ask for more NO get it into your head overclocking kills JUST LIKE SMOKING gosh you need to set the stock defaults and if you want to oc then watch many tutorials (the detailed ones which explain why you do one thing to have an effect on another thing , i mean im sure you dont need to oc (its a FUCKING GTX 970) secondly you dont need to im sure you will regret damage to your card. im sorry for being so hard on you but so many people nowerdays break there cards because they follow a crummy youtube video which does not explain anything to be good at computers you must MUST know your computer and the limit of each component

Sir, you need to calm your language and attitude. Thanks for spreading the understanding but you make OC seem like some life threatening situation. :lol:

 

[amd4lif3]

I'm calm now but it seams like some people just want to OC until the cards dead its important to have limits thats all im saying and to me graphics cards are people too XD

 

[Dunkan77]

I'm calm now but it seams like some people just want to OC until the cards dead its important to have limits thats all im saying and to me graphics cards are people too XD

I know that it kills it and it's why I DO NOT play at such OC's. I just want to see once,just once how far I can push it and I only run it a those speeds for 1 or 2 mins just to test stability. But does OV without excess heating damage it as well? And do Backplates help cool things down a bit like the VRM and memory chips?

 

[amd4lif3]

basically when you increase the voltage the heat increases not to mention the fact that the components are meant to run what they are stock or maybe a bit higher , but noone should ever try to reach the limit of a chip it may encounter an error because of this its like trying to walk on thin ice possible to reach the limits but not reccomended and vrm s are cooled by the backplate sometimes there are thermal pads also

 

[Dunkan77]

basically when you increase the voltage the heat increases not to mention the fact that the components are meant to run what they are stock or maybe a bit higher , but noone should ever try to reach the limit of a chip it may encounter an error because of this its like trying to walk on thin ice possible to reach the limits but not reccomended and vrm s are cooled by the backplate sometimes there are thermal pads also

So it is better although still dangerous with a backplate? I will be getting one anyway because it looks better and the memory chips have thermal pads for them. Thank you for helping 😀

 

[Dunkan77]

It can either be the GPU or the monitor. Try using another display to confirm what exactly is producing this issue.

If it is the GPU, what PSU are you using?

And artifacting doesn't usually damage the monitor.

Sorry I didn't answer, I have a Corsair RM650 80+Gold 650 W PSU so it's more than enough for this system as max power ever drawn was when I used GPU and CPU burners at the same time and was about 300-400 W

 

[amd4lif3]

yes that psu is good enough

 

[amd4lif3]

basically when you increase the voltage the heat increases not to mention the fact that the components are meant to run what they are stock or maybe a bit higher , but noone should ever try to reach the limit of a chip it may encounter an error because of this its like trying to walk on thin ice possible to reach the limits but not reccomended and vrm s are cooled by the backplate sometimes there are thermal pads also

So it is better although still dangerous with a backplate? I will be getting one anyway because it looks better and the memory chips have thermal pads for them. Thank you for helping 😀

overclocking is safe unless you want to reach the limits since when doing this your components have no room or error so i dont reccomend ocing to the limits , good i like a gpu with a backplate it looks soo darn good , and no problem