I’m currently playing 3 games including WOW, and I often got pixel screen error which forces my pc to restart whenever i play one of them. Two of them including WOW , i don’t have pixel screen error no matter how long i play them. But for that one game, I usually got pixel screen error in a short time. Is it my GPU error or something else? I’m using 3080 TI graphic card.
1 Like
What is the one game you’re having trouble with?
Did you make sure everything was up to date? It could be a simple with by having your drivers and windows up to date.
is it a blue error screen?
Which “one game” is actually causing the problem? You’ve told us what it isn’t… partially… but we have no idea what it is. Everything puts a different load on the GPU - WoW, for example, is very heavy on pixel throughput, which in turn leans heavily on the raster side of things and bus coherency. Without knowing what load your hardware is struggling with no real diagnostics can be done, and the possibility exists that it’s simply faulty software as well.
Your description is also extremely vague, and I suspect not even accurate. “Pixel screen error” doesn’t actually describe anything meaningful. I’m sure to you it does, but to anyone else they’re going to have to guess what you’re talking about - which is a really terrible way to perform diagnostics.
I’m playing 3 games atm which are Dota 2, CS GO and WOW. Dota 2 is the only game i have pixel screen error in a short period of time. I mean when i play CS GO and WOW, it seems i don’t have any problem at all. At first i thought my new GPU card is dead, but it turns out when i play WOW and CSGO, there isn’t any problem at all But i’m having error which is when i play dota 2, i might get green, red pixels all over my monitor and it forced my pc to restart itself.
It’s not a blue screen error, it’s like dead pixels appearing on my monitor.
Discoloured pixels suggests perhaps a vRAM issue. If you’re overclocking your vRAM - at all - set it back to factory. I don’t care if it’s one of those “all 3080s can do this speed” things people spout online, set it to factory because they can’t. Corruption in vRAM could easily trigger a restart due to modern GPUs being programmable, where older, fixed-function cards could carry on while simply displaying corruption.
You could also try lowering the load on your vRAM. Impose a framerate limit through the control panel or enable v-sync (if you don’t already have it enabled). This will unfortunately have the knock-on effect of lowering GPU load as well so it’s not an isolated test, but it will at least let you know if load is an issue or not. Note that adjusting resolution won’t have the same effect as frame rates don’t drop as fast as pixel counts increase.
It may also be worth investigating your vRAM temperatures, using a tool like GPU-Z. Not all cards will display this (my 3070 doesn’t), unfortunately. While the memory is rated to 110°c, anything over 90°c could indicate poor contact between the vRAM and the cooling. Various hardware YouTube personalities (eg JayzTwoCents) have demonstrated replacing vRAM cooling interface on cards like the 3080s to sometimes significant effect.
yeah as the previous poster said, take off any overclock. also if we can get the full spec of the PC, especially the power supply
1 Like
That’s a fair concern I hadn’t considered. NVidia’s 30 series in particular are known to exceed their specified TDP - sometimes quite significantly. I haven’t seen any explicit testing on the 3080Ti but the 3090 has been shown going over 100w past its TDP (350w) without any user intervention, such as raising the power limit.
A 3080 TI will scoff at most things wow can throw at it. The hardware itself is fine. Maybe update drivers?