Gpu problems

[Mythbredor]

Corsair Cx500.

Edit: I originally had a cooler master PSU (can't remember exact one) but was a discontinued line, so couldn't straight replace.

The CX line in corsair is known to have some bad components sometimes, I would suggest replacing your PSU first or testing your GPU in a friends system.

[Darsithis]

I only recently replaced my PSU in May. So I'm bloody hoping this isn't the problem!

You'll probably pay less for a replacement PSU than a 1080 GTX.

[Deleted]

I updated my drivers. Didn't fix anything.

So is the consensus that the PSU is the problem? If so what model would you guys then recommend?

Edit: screens went black. Not coming back and PC isn't rebooting, first time it's gone like this.

[Sturmbringe]

If you get arties all over the screen, 99% it''s your GPU and not the PSU.

Corsair CX is an incredibly bad PSU rated Tier 4:"Built down to a low price. Not exactly the most stable units ever created. Very basic safety circuitry or even thin gauge wiring used. Not for gaming rigs or overclocking systems of any kind. Avoid unless your budget dictates your choice"

http://www.tomshardware.co.uk/forum/...tier-list.html

However, it doesn't sound like your PSU is to blame here.

[Mythbredor]

What Strumbringe said could be true but next best step is testing your GPU in another system.

[Deleted]

What Strumbringe said could be true but next best step is testing your GPU in another system.

And if I'm unable to do that what's my next best plan f action?

- - - Updated - - -

Update: Might have found a computer I can test the GPU in. If it does turn out to be the PSU any recommendations for which one to get as a replacement?

[Evildeffy]

Are you still getting green lines/artefacts?

Or is it just black now?

Because if it's going full black and nothing .. then there's a distinct possibility it's the PSU.
Having said that ... 32°C on the GPU even in idle is a bit awkward, that's extremely low.. are you sure you've seen that right?

Can you provide a screenshot of that?

If there's green lines/artefacts happening and the fans are spinning and almost no build-up of dust in the graphics card heatsink then it's VERY likely the card is dying.
Can you remember any other detail that it started doing or did to this?

[Deleted]

[QUOTE=Strifus;43690766] I've got an Amd Radeon HD 7800 series?



Thats you're problem right there. I had exactly the same issues and decided to upgrade my gpu on a whim (lending my friend the old gpu and his pc crashed non stop after installing) Since that point my pc has been fine

[Sturmbringe]

And if I'm unable to do that what's my next best plan f action?

- - - Updated - - -

Update: Might have found a computer I can test the GPU in. If it does turn out to be the PSU any recommendations for which one to get as a replacement?

IF the PSU turns out to be the faulty component, then I recommend buying any of the Tier 3 or Tier 2 PSU's as noted below:

http://www.tomshardware.co.uk/forum/...tier-list.html

I still think your GPU is failing. This is how to test: Download 3D Mark 11 and 3D mark Fire Strike, lower your GPU's clocks to stock, and then run both benchmarks at max settings (1920X1080 for the first and 2560X1440 for the second).

[Deleted]

Are you still getting green lines/artefacts?

Or is it just black now?

Because if it's going full black and nothing .. then there's a distinct possibility it's the PSU.
Having said that ... 32°C on the GPU even in idle is a bit awkward, that's extremely low.. are you sure you've seen that right?

Can you provide a screenshot of that?

If there's green lines/artefacts happening and the fans are spinning and almost no build-up of dust in the graphics card heatsink then it's VERY likely the card is dying.
Can you remember any other detail that it started doing or did to this?

Still getting artifacts and then it went black a good few minutes afterwards.

Definitely read it correctly as 32 degrees. Whether that temperature is actually correct or not I don't know, but it definitely displayed 32.

The first sign something was going even slightly wrong was that the game I was playing, Grim Dawn, froze a couple of times, or would go black if I tabbed out and then back in. I then rebooted the game again, loaded into the world and there was a black triangle shape stretching across the screen but it was static in the game world, as in I could move underneath and past it, as if the textures hadn't loaded. The game then froze and shortly after artifacts started to appear. Prior to this the game and everything else had been running. Perfectly fine with no problems.

[Evildeffy]

Still getting artifacts and then it went black a good few minutes afterwards.

Definitely read it correctly as 32 degrees. Whether that temperature is actually correct or not I don't know, but it definitely displayed 32.

The first sign something was going even slightly wrong was that the game I was playing, Grim Dawn, froze a couple of times, or would go black if I tabbed out and then back in. I then rebooted the game again, loaded into the world and there was a black triangle shape stretching across the screen but it was static in the game world, as in I could move underneath and past it, as if the textures hadn't loaded. The game then froze and shortly after artifacts started to appear. Prior to this the game and everything else had been running. Perfectly fine with no problems.

Artefacts that show as if it hasn't loaded on an object is memory corruption which can be due to power, card or software issues.
Artefacts that are blocks/tears and are consistently moving along with you and not on ingame objects are due to possible overheating or a dying card.
(What I mean by this is like they are more attached to your monitor than game)

Can you define which of the 2 it is?

Also can you post a screenshot using HWMonitor from http://www.cpuid.com for the graphics as that 32°C is irking in the back of my mind (the wrong way).

[Deleted]

Artefacts that show as if it hasn't loaded on an object is memory corruption which can be due to power, card or software issues.
Artefacts that are blocks/tears and are consistently moving along with you and not on ingame objects are due to possible overheating or a dying card.
(What I mean by this is like they are more attached to your monitor than game)

Can you define which of the 2 it is?

Also can you post a screenshot using HWMonitor from http://www.cpuid.com for the graphics as that 32°C is irking in the back of my mind (the wrong way).

So there was that one instance of the black object on my screen in game. But I'm also getting blocks and lines on my desktop that react and change to my mouse movemen even when I have no programmes open.

I can't currently screenshot it as I've now gone to bed as my partner is ill. Earliest I can do that will be tomorrow morning. But I'm 100% certain it said 32.0 degrees, it doesn't seem right but to me either as it never changed from 32.0 at all.

[Evildeffy]

So there was that one instance of the black object on my screen in game. But I'm also getting blocks and lines on my desktop that react and change to my mouse movemen even when I have no programmes open.

If they react to movement on screen but not on Icons f.ex. then it's your card.
What I mean by this is that when you bring an icon over that green artefact area and you drop the icon there that it doesn't look like you used an eraser in MS Paint but it more looks like it's fully independent of your desktop.

I can't currently screenshot it as I've now gone to bed as my partner is ill. Earliest I can do that will be tomorrow morning. But I'm 100% certain it said 32.0 degrees, it doesn't seem right but to me either as it never changed from 32.0 at all.

That's why I asked if you could do so with the aforementioned programme, it is more reliable in terms of catching sensors and temperatures.
But if you can with your smartphone make a clear picture of the green artefacts on your screen, that'd be helpful as well.

But it does sound more and more like graphics card dying in any case.

[Remilia]

Artifacts are caused mainly by a few things and very rarely connected to the PSU, while possible, it's unlikely.

Typical artifacts can be not enough voltage for the frequency or consequently too high frequency for the voltage. This can be an issue by age or by overclocking too high after a while. The other possibility is VRAM is defective or aged too much, which at that point you can't do anything.

You can try something like MSI afterburner and lower the frequency a bit and see how it goes. Easier than replacing it if you don't need to.

If the PSU was an issue, more than your graphics card is going to be affected.

[Deleted]

If all else fails.

[Evildeffy]

Artifacts are caused mainly by a few things and very rarely connected to the PSU, while possible, it's unlikely.

Typical artifacts can be not enough voltage for the frequency or consequently too high frequency for the voltage. This can be an issue by age or by overclocking too high after a while. The other possibility is VRAM is defective or aged too much, which at that point you can't do anything.

You can try something like MSI afterburner and lower the frequency a bit and see how it goes. Easier than replacing it if you don't need to.

If the PSU was an issue, more than your graphics card is going to be affected.

Not entirely.
I've had such things happen exactly as described because of PSU issues, f.ex. a malfunctioning 12V rail etc. where the other rails work fine.

Since the CX-series aren't too well received this is entirely in the realm of possibility.

However as described in the sense before if it is the graphics card then your suggestion is only a band-aid fix, not entirely a solution but it could extend it's life a bit longer.. however the OP has already given thought of replacement.. we're just making sure it is what it is.

[Deleted]

If they react to movement on screen but not on Icons f.ex. then it's your card.
What I mean by this is that when you bring an icon over that green artefact area and you drop the icon there that it doesn't look like you used an eraser in MS Paint but it more looks like it's fully independent of your desktop.

The artifacts are almost like an overlay, it's more like they're appearing on top of my desktop than actually being a part of it. If that helps at all. They're also not just green, very multicoloured, generally square in shape with some lines appearing as well. Sometimes it creates and almost chess board like image with white artifacts. There is essentially no consistency in them at all.

I apologise I can't get screenshots at the moment, best I can do is provide as much information as I can.

I also did download a hardware monitor which was giving me the 32 reading, not a built in piece of software or anything.

[Nakotsu]

I'm not entirely sure of my exact specifications as my computer keeps freezing and leaving artifacts. I do know I've got an Amd Radeon but unsure on the specific model, computer is around 3 years old.

Anyway, my (what I'm assuming) GPU has started to die. Last night multiple games started to crash and there would be artifacts all over the screen. My PC's core temp was only around 22 celsius but I felt the GPU and it was burning to the touch.

I shut down my PC thinking it might just be overheating and I'll let it cool and try tomorrow. Fast forward to about 20 minutes ago, I boot up my PC, it's fine, doing some stuff deleting old files etc. I boot up a game and straight away it crashes and artifacts. My screen then went striped with grey lines and my PC rebooted itself.

My general question is; what the feck is going on? Has my GPU died? Am in need of a replacement?

Thank you for the help.

Ive identifiyed your problem get a nvidia card instead.

you will either A have to format the system as it either has a software issue or B a power issue might be ur power supply died could be the video card itself...

[Remilia]

Not entirely.
I've had such things happen exactly as described because of PSU issues, f.ex. a malfunctioning 12V rail etc. where the other rails work fine.

Since the CX-series aren't too well received this is entirely in the realm of possibility.

However as described in the sense before if it is the graphics card then your suggestion is only a band-aid fix, not entirely a solution but it could extend it's life a bit longer.. however the OP has already given thought of replacement.. we're just making sure it is what it is.

Oh yes it's definitely a bandaid solution but if you can squeeze another year out with no problems then I'd do it, that's just me though.

The thing is for PSU though is if the 12V rail is bad, then your CPU and all will be affected. The only thing that runs on the 3.3V and 5V are peripherals and storage drives at this point. CPUs no longer run on the 5V lane.

[Lathais]

The CX line in corsair is known to have some bad components sometimes, I would suggest replacing your PSU first or testing your GPU in a friends system.

They used to, but they remedied the issues. Still not a top quality PSU, but the newer ones aren't bad. He said he just got it in May. It's most likely a graphics card problem.

- - - Updated - - -

Are you still getting green lines/artefacts?

Or is it just black now?

Because if it's going full black and nothing .. then there's a distinct possibility it's the PSU.
Having said that ... 32°C on the GPU even in idle is a bit awkward, that's extremely low.. are you sure you've seen that right?

Can you provide a screenshot of that?

If there's green lines/artefacts happening and the fans are spinning and almost no build-up of dust in the graphics card heatsink then it's VERY likely the card is dying.
Can you remember any other detail that it started doing or did to this?

32C is low? Mine is currently running at 28C. I know those AMD cards run hotter, but mine is also only a single fan, his is dual. Also, I used to have a 6870, which would run pretty hot, but would idle in the low 30s like that.

- - - Updated - - -

IF the PSU turns out to be the faulty component, then I recommend buying any of the Tier 3 or Tier 2 PSU's as noted below:

http://www.tomshardware.co.uk/forum/...tier-list.html

I still think your GPU is failing. This is how to test: Download 3D Mark 11 and 3D mark Fire Strike, lower your GPU's clocks to stock, and then run both benchmarks at max settings (1920X1080 for the first and 2560X1440 for the second).

Oh god, not the tier list that has literally the exact same PSU just with a different label on it in two different tiers. That thing is not to be trusted.