Strange variation on the AMD 'vertical lines crash'

[Sergeant Dornan]

In the very likely case that a lot of readers don't know what the crash is about, during games or otherwise, users of AMD GPUs will sometimes find their PCs' displays replaced by a bunch of vertical coloured lines. The PC would freeze soon afterwards. This has been happening to me for some time now as well. I've tried fiddling around with AMD Overdrive, turning it off and on and adjusting memory clocks (which would only lengthen the period of time the PC would go without such a crash). I have done a couple of stress tests on the CPU which is an AMD 6300 hexa-core, the GPU, an AMD HD 7870, and a probably unimportant HDD while monitoring all temperatures, but everything appeared to be proper.
Thing is, the crashes occur ONLY while I'm in certain games. Before you think it's related to temperature, I can give an example. While I'm in TF2, for instance, the GPU might reach 56 degrees and the PC would crash after only having played it for half an hour, whereas while I'm in Civilization V it easily reaches 62+ degrees while having no problem with prolonged stays of several hours. I thought it would be an engine thing, which it probably isn't. Where it crashes in Half Life 2 and Red Orchestra 2, it won't do anything in EYE or BioShock Infinite.
The GPU is still within warranty, but I would appreciate any advice given before I have to take that step.

[_I_]

what board?

fx6300 has a max of 62-72c
if the cpu hits 62c it could be thermal throttling
or if its a weak board it could be vrm throttling thats causing the crash ect..

and what power supply (brand/model)
poor psu can cause crash/bsod/reboot ect..

[Sergeant Dornan]

For a motherboard I have a Gigabyte 970A-DS3 and the power supply is a Thermaltake Smart SE 530.

[Azza ☠]

If I remember rightly - Horizontal lines relate to the monitor, as that is the direction it draw processes. While vertical lines relate to the graphics card. Or it might of be visa versa? I can't remember which is which to be honest off the top of my head right now.

Anyways, the first thing to check and cross your fingers, is just a cable fault. However, that won't cause any system crashing, nor would it be if just the monitor dying.

If you use any kind of cable spitter or converter, get rid of it. They suck, lose quality, and break regularly over usage. VGA-to-DVI converters are the worst, I personally set them on fire. Anyways, check around the ports on both the graphic card and the monitor end.

If not, it's most logically then the graphic card damaged, overheating, or a driver issue. If the video memory is overheating/damaged artifacts might start appearing on screen (stretching of particles across the screen, lines, dots, blocks, etc). It will corrupt and get worst over time. The hardware might need to be tested and replaced in that case.

[Sergeant Dornan]

I was thinking that the graphics card would be damaged, too, but then if it would be so, it would definitely happen more often, without any regard of the game I was in. The main thing that leaves me confused is why it happens in some games, but not in others.

[_I_]

get a better power supply

thermaltake low end are very poor

amd build = 300w
7870 = 170w

way too much for the tt 530w to run and be stable

[Sergeant Dornan]

Somehow I don't think that's true, because if it were, the power supply would have already failed while I was doing the stress blend test (FurMark and Prime95 at the same time, and nothing happened in any way).

[_I_]

because it hasnt killed any components yet doest mean it wont