Issues with graphics card after 4 RMA's

[jamiewhite]

I've had four RMA's from EVGA, after my first 580 failed I was upgraded to a 680 which was swapped with another 680 and now this most recent RMA I was upgraded to a 770 (Yes EVGA are that awesome).

Unfortunately, I'm still having problems. The most recent issue occurred for no reason at all one day when I booted my PC up. It booted in an ultra low resolution and low colour mode. After a few hours of troublshooting my graphics drivers, I found that installing the standard Microsoft Windows drivers got into a non-low colour mode which was a bit easier on the eyes.

The problem is that as long as the PC is powered on, from BIOS to desktop, there are strange colourful artefacts moving across or static on the screen. They are red and blue and some stay in the same place in clusters or several singular ones moves across the screen.


A rather telling thing is that screenshotting whilst on the PC reveals non of these issues (except of course the low resolution and low colour mode with the Nvidia drivers). This indicates to me these artefacts are only manifesting themselves on the monitor itself. I can only assume that it is a motherboard problem, others have suggested PSU issues.

A photo of the screen artefacts:

The screenshots (minus the screen artefacts because as above) -

My desktop with Nvidia drivers:

My desktop with standard Microsoft Windows drivers:

Has anyone had any experience with these issues previously?
Does anyone know what is wrong?


After 9 months of issues and 4 new graphics card I'm going slightly insane -_-



UPDATE:
Completely forgot to post my specs >_<


CPU: Intel Core i5-2500K

CPU COOLER: Cooler Master Hyper 212 EVO

MOBO: Gigabyte GA-Z68XP-UD3 (Rev. 1.0)

RAM: Corsair Vengeance Low Profile 8GB (2x4GB) DDR3 1600 Mhz CL9

GPU: EVGA GTX 770

PSU: Corsair TX750M 750 watt

PERIPHERALS: 2 x Acer S220HQLBrbd 21.5 inch

 

[fjaesbog]

There must be something else than the graphics card causing this issue. 4 GPU cards through 9 months isn't really normal though.
Is there by any chance another PSU / Monitor you could test your GPU on? This could also be your monitor causing artifacts, not only the GPU.

 

[Bee_Dee_3_Dee]

There must be something else than the graphics card causing this issue. 4 GPU cards through 9 months isn't really normal though.
Is there by any chance another PSU / Monitor you could test your GPU on? This could also be your monitor causing artifacts, not only the GPU.

yes, rule out hardware.

maybe software wise...

i read this thread last nite. and after thinking a while, i figured 2D versus 3D (Legacy/ Microsoft drivers versus nVidia drivers) was maybe a good clue.

and therefore, the temperature of the GPU could be causing the anomalies when running the vid card with nVidia drivers versus legacy drivers. (just think... when ur OCing a vid card... u throttle bak, when anomalies appear. right?)

and with Legacy drivers, the GPU is not pushed as much; because the CPU does much more. and the temps stay down.

so, ^^ is a guess. but if jamiewhite checks the temps of the GPU with a utility he can report the temps.

until the temps are confirmed normal or high, there's no need to explain further.

But if temps are too high, the solution is to adjust the speed of the GPU's fan by using MSI Afterburner. AND (first) proper uninstall reinstall of vid drivers too. Because the OS could be very confused after so many different vid cards, right? and the poorly calibrated fan speeds that most cards come with could be made worse as a result of vid drivers not being properly uninstalled and reinstalled.

 

[hang-the-9]

I've had four RMA's from EVGA, after my first 580 failed I was upgraded to a 680 which was swapped with another 680 and now this most recent RMA I was upgraded to a 770 (Yes EVGA are that awesome).

Unfortunately, I'm still having problems. The most recent issue occurred for no reason at all one day when I booted my PC up. It booted in an ultra low resolution and low colour mode. After a few hours of troublshooting my graphics drivers, I found that installing the standard Microsoft Windows drivers got into a non-low colour mode which was a bit easier on the eyes.

The problem is that as long as the PC is powered on, from BIOS to desktop, there are strange colourful artefacts moving across or static on the screen. They are red and blue and some stay in the same place in clusters or several singular ones moves across the screen.


A rather telling thing is that screenshotting whilst on the PC reveals non of these issues (except of course the low resolution and low colour mode with the Nvidia drivers). This indicates to me these artefacts are only manifesting themselves on the monitor itself. I can only assume that it is a motherboard problem, others have suggested PSU issues.

A photo of the screen artefacts:

The screenshots (minus the screen artefacts because as above) -

My desktop with Nvidia drivers:

My desktop with standard Microsoft Windows drivers:

Has anyone had any experience with these issues previously?
Does anyone know what is wrong?


After 9 months of issues and 4 new graphics card I'm going slightly insane -_-

Something in your system is killing the video cards. You could be overclocking too much, you have too much heat, your power supply is not good enough to run the cards. Those 3 cover probably 90% of what could be causing your issues.

 

[jamiewhite]

There must be something else than the graphics card causing this issue. 4 GPU cards through 9 months isn't really normal though.
Is there by any chance another PSU / Monitor you could test your GPU on? This could also be your monitor causing artifacts, not only the GPU.

The thing is that only the first of these cards showed any actual errors when tested by EVGA (the manufacturer) themselves. As far as they are concerned I've sent back two working cards and scammed them out of three cards till I got upgraded from a 580 to 770. It looks terrible right now.
I've tested both my monitors on another PC and tried each one on it's own in my PC with the same results.
As to the PSU, I don't have a spare one.
I do suspect both the PSU and motherboard to be the root of this issue but I don't know how to narrow it down or test without a myriad of spare parts 🙁


P.S. I updated the original post with my specs Completely forgot.

 

[Bee_Dee_3_Dee]

but wat is the GPU temperature when u 1st start the PC? (after it is off at least 15 minuets.) and wat are the GPU temps after just sitting running nothing but the desktop for 15 minuets? let alone, trying to start a vid game or playing a few minuets?

 

[jamiewhite]

yes, rule out hardware.

maybe software wise...

i read this thread last nite. and after thinking a while, i figured 2D versus 3D (Legacy/ Microsoft drivers versus nVidia drivers) was maybe a good clue.

and therefore, the temperature of the GPU could be causing the anomalies when running the vid card with nVidia drivers versus legacy drivers. (just think... when ur OCing a vid card... u throttle bak, when anomalies appear. right?)

and with Legacy drivers, the GPU is not pushed as much; because the CPU does much more. and the temps stay down.

so, ^^ is a guess. but if jamiewhite checks the temps of the GPU with a utility he can report the temps.

until the temps are confirmed normal or high, there's no need to explain further.

But if temps are too high, the solution is to adjust the speed of the GPU's fan by using MSI Afterburner. AND (first) proper uninstall reinstall of vid drivers too. Because the OS could be very confused after so many different vid cards, right? and the poorly calibrated fan speeds that most cards come with could be made worse as a result of vid drivers not being properly uninstalled and reinstalled.

I'm not sure how it could be GPU temps as I had MSI Afterburner and Rivatuner installed and running the On-Screen display all the time and I never broke 60°C.
As I mentioned above as well, I have sent 3 graphics cards back and upon testing at the manufacturers hands I've been told that only the first card I sent back (my original GTX 580) had a problem and that was that one of the DVI ports stopped working.
The rest have been fully tested and EVGA (my manufacturer) thing I'm crazy >_< or trying to get upgrades. The answer is, this issue clearly lies elsewhere than the graphics card.

Unfortunately, MSI Afterburner won't post any information anymore, all the readings are greyed out. I'll try again with the Nvidia drivers installed but otherwise I cannot get temperature readings. The PC itself is unusable and is never any longer left on for more than a minute (to take screenshots and test things).

As to the OS and all the drivers, I bought an SSD before this card got here and reinstalled Windows fresh so the drivers aren't the problem. I uninstalled them and reinstalled just in case by unhiding them in device manager and completely cleaning them but I still get the same artefacts and low colour mode. Only the standard Windows drivers help with that, bringing back full colour mode.
I also have a custom fan curve.

I am at a loss as to what it could be, but I am in touch with Corsair (PSU) and Gigabyte (motherboard) but when they get back to me and how helpful they are is yet to be determined.

P.S. I updated the original post with my specs Completely forgot.

 

[jamiewhite]

Something in your system is killing the video cards. You could be overclocking too much, you have too much heat, your power supply is not good enough to run the cards. Those 3 cover probably 90% of what could be causing your issues.

As mentioned above I have MSI Afterburner and Rivatuner installed with a custom fan curve and the on-screen display on at all times. I haven't yet seen heats over 60°C. My power supply is a 750w 80 Plus Bronze Corsair TX750M.
Everything you said is correct but I think that my PSU and temp monitoring is sufficient enough to rule this out, however I do think that perhaps I have a faulty PSU or motherboard?


P.S. I updated the original post with my specs Completely forgot.

 

[Bee_Dee_3_Dee]

Something in your system is killing the video cards. You could be overclocking too much, you have too much heat, your power supply is not good enough to run the cards. Those 3 cover probably 90% of what could be causing your issues.

As mentioned above I have MSI Afterburner and Rivatuner installed with a custom fan curve and the on-screen display on at all times. I haven't yet seen heats over 60°C. My power supply is a 750w 80 Plus Bronze Corsair TX750M.
Everything you said is correct but I think that my PSU and temp monitoring is sufficient enough to rule this out, however I do think that perhaps I have a faulty PSU or motherboard?


P.S. I updated the original post with my specs Completely forgot.

i see no mention of MSI Afterburner and Rivatuner. but np ur no nub and that's all that matters. 😉

only prob is the diag is getting more complicated.

list all u tried so far and use numbers.

and to rule out all system drivers; much of the time a manufacturer of hardware will insist on trying a new freshly installed OS. but u can start with saying if u have the latest Intel Chipset drivers installed for the MOBO. Or are they all legacy?

 

[jamiewhite]

i see no mention of MSI Afterburner and Rivatuner. but np ur no nub and that's all that matters. 😉

only prob is the diag is getting more complicated.

list all u tried so far and use numbers.

and to rule out all system drivers; much of the time a manufacturer of hardware will insist on trying a new freshly installed OS. but u can start with saying if u have the latest Intel Chipset drivers installed for the MOBO. Or are they all legacy?

Oh no, I didn't mean that in a snarky way! I was just referring to my replies to other people (which I forgot would be below that one >_< doh!). I didn't mean to cause any offence, it's just my wording 🙁
I'm going to test the things you've said and write a clearer response.

The problem with a freshly installed OS is that these issues (the coloured artefacts) present themselves on the Windows installation disk. From BIOS to even installing a new OS the strange colours in the first screenshot cover the screen. This indicates to me it's not a driver or software issue but a hardware problem.

I'll get back to you on temps and the testing that you mentioned though.

 

[Bee_Dee_3_Dee]

i see no mention of MSI Afterburner and Rivatuner. but np ur no nub and that's all that matters. 😉

only prob is the diag is getting more complicated.

list all u tried so far and use numbers.

and to rule out all system drivers; much of the time a manufacturer of hardware will insist on trying a new freshly installed OS. but u can start with saying if u have the latest Intel Chipset drivers installed for the MOBO. Or are they all legacy?

Oh no, I didn't mean that in a snarky way! I was just referring to my replies to other people (which I forgot would be below that one >_< doh!). I didn't mean to cause any offence, it's just my wording 🙁
I'm going to test the things you've said and write a clearer response.

The problem with a freshly installed OS is that these issues (the coloured artefacts) present themselves on the Windows installation disk. From BIOS to even installing a new OS the strange colours in the first screenshot cover the screen. This indicates to me it's not a driver or software issue but a hardware problem.

I'll get back to you on temps and the testing that you mentioned though.

no snarks detected.

if it (anomalies) happens during OS installs too; there's more obvious questions.

but just one for now...

have u tried the 8x PCI-E slot?

 

[jamiewhite]

no snarks detected.

if it (anomalies) happens during OS installs too; there's more obvious questions.

but just one for now...

have u tried the 8x PCI-E slot?

Good because I in no way meant to offend. I'm still a nub cake but unlike a lot of people that aren't educated or work in this area I take a lot of time to research and learn.
That's kind of a problem with these manufacturers, they assume you know nothing and treat you like so which just delays the process. It's why I came here instead, this place is freaking awesome for help and advice. I love it

As to my situation, I've booted the PC and run it with and without the nvidia drivers to check temps both at boot and after about 15 minutes of idling -
Standard Windows drivers: 27°C
Nvidia official drivers: 30°C
And they never really fluctuate between a few degrees from there after keeping the pc on for a while.

Unfortunately I can't run a game because they all crash on startup. They don't even open. Steam doesn't start and games just post errors without even loading up the main menu.

I have also tested both PCIE slots under the instruction of Gigabyte and both present the same issues.

So temps are ridiculously low no matter what and have never been better as this 770 is the one with ACX cooling.
No drivers work/change the situation no matter how clean an install I do.
Both PCIE slots have the same issues.
And even sitting in the bios has these artefacts on the screen.
And the artefacts are only showing on the monitor, not when I take screenshots.

One of my friends mentioned it could be the PSU not serving enough power or a broken motherboard.

This has been happening since December 2013 and I've gone through 4 graphics cards, all whilst trying to get a streaming channel going on Twitch

 

[mlga91]

Try starting into safe mode and sweep all nvidia drivers:
http://www.guru3d.com/content-page/guru3d-driver-sweeper.html

And then perform a clean driver install, dont install the same ones that are giving you problem.

 

[jamiewhite]

Try starting into safe mode and sweep all nvidia drivers:
http://www.guru3d.com/content-page/guru3d-driver-sweeper.html

And then perform a clean driver install, dont install the same ones that are giving you problem.

I did try this, I first used DDU (Display Driver Uninstaller) and then when that didn't work I used the command line to unhide all drivers that my partners brother who's a network admin showed me and then manually uninstalled them from there and reinstalled. He ran me through quite a few tests and determined it wasn't the drivers. Especially since this happens when in the BIOS :/

 

[Bee_Dee_3_Dee]

Try starting into safe mode and sweep all nvidia drivers:
http://www.guru3d.com/content-page/guru3d-driver-sweeper.html

And then perform a clean driver install, dont install the same ones that are giving you problem.

I did try this, I first used DDU (Display Driver Uninstaller) and then when that didn't work I used the command line to unhide all drivers that my partners brother who's a network admin showed me and then manually uninstalled them from there and reinstalled. He ran me through quite a few tests and determined it wasn't the drivers. Especially since this happens when in the BIOS :/

it's starting to look like u have tried many things it's hard to keep track of. (this is good.) but try making a list (so it is more apparent what is left to do) similar to the following:


Problem: "The problem is that as long as the PC is powered on, from BIOS to desktop, there are strange colourful artefacts moving across or static on the screen."
Tried:
___ Clean uninstall reinstall of drivers
___ RMAed video card (4xs)
___ Checked BIOS/ reset
___ Fresh OS installation
___ Legacy drivers seem to work fine (But everyone know u can't play video games with Legacy drivers.)
___ "The thing is that only the first of these cards showed any actual errors when tested by EVGA (the manufacturer) themselves."
___ GPU Temps are fine ("never broke 60°C.")
___ Tested Monitors on a different PC. [Did they worked fine, Jamiewhite?]
___ PSU not tested yet ("As to the PSU, I don't have a spare one.")
___ Checked device manager for hidden "Ghost" drivers
___
___
___
___

Go ahead and add to ^^.

So far, with ^^, a different PSU is the 1st must do.
And updating the BIOS sticks out.
But were each of the 4 vid cards, tested on a working PC or not?
Unplugging all unnecessary peripherals is a golden rule.
Are u using a UPS (Uninterruptible Power Supply)? (this has been overlooked before by ppl for months.)
Speaking of UPS a short/ connection on the MOBO sounds very likely.
Remove the MOBO from the case and place mobo box on top of the anti-static wrap. Start it by shorting the Pins with screwdriver. (this should be done before RMAing MOBO.)

All kinds of things could still be questioned. but u might as well complete everything done/ Tried already list ^^ first.

GL

 

[junkeymonkey]

I wonder if the first card that evga proved to have a issue if that issue did damage to your board in some way and now your screwed all around ???

 

[jamiewhite]

it's starting to look like u have tried many things it's hard to keep track of. (this is good.) but try making a list (so it is more apparent what is left to do) similar to the following:


Problem: "The problem is that as long as the PC is powered on, from BIOS to desktop, there are strange colourful artefacts moving across or static on the screen."
Tried:
___ Clean uninstall reinstall of drivers
___ RMAed video card (4xs)
___ Checked BIOS/ reset
___ Fresh OS installation
___ Legacy drivers seem to work fine (But everyone know u can't play video games with Legacy drivers.)
___ "The thing is that only the first of these cards showed any actual errors when tested by EVGA (the manufacturer) themselves."
___ GPU Temps are fine ("never broke 60°C.")
___ Tested Monitors on a different PC. [Did they worked fine, Jamiewhite?]
___ PSU not tested yet ("As to the PSU, I don't have a spare one.")
___ Checked device manager for hidden "Ghost" drivers
___
___
___
___

Go ahead and add to ^^.

So far, with ^^, a different PSU is the 1st must do.
And updating the BIOS sticks out.
But were each of the 4 vid cards, tested on a working PC or not?
Unplugging all unnecessary peripherals is a golden rule.
Are u using a UPS (Uninterruptible Power Supply)? (this has been overlooked before by ppl for months.)
Speaking of UPS a short/ connection on the MOBO sounds very likely.
Remove the MOBO from the case and place mobo box on top of the anti-static wrap. Start it by shorting the Pins with screwdriver. (this should be done before RMAing MOBO.)

All kinds of things could still be questioned. but u might as well complete everything done/ Tried already list ^^ first.

GL

You pretty much got everything there in your list. The only thing I would add is that I was on a Gigabyte UEFI BIOS and under their instruction I flashed back to the old Award BIOS which unfortunately didn't fix anything.

As to your points, I can try and rustle up an extra PSU from friends and see if I can test that at the weekend. Though Corsair themselves seem pretty adamant that it isn't a PSU problem :/
So the BIOS has been reflashed to a UEFI one and then downgraded to Award and neither worked.
All 4 GPU's were working for a time in my rig but then after a few weeks/months they all started to show issues. Both 680's were tested by EVGA in their factory and I was told their were no issues whatsoever.
All peripherals were unplugged pretty much straight away 😉
I don't have an UPS unfortunately and couldn't afford one 🙁 But I have tested the PC and monitor both plugged straight into the wall socket since and still the same issues.
When I put the mobo on top of the mobo box do you want me to plug the monitor/storage in and boot it up like that?

UPDATE:
Now after reflashing that award BIOS and rebooting again my PC no longer boots and goes straight to a completely black screen with a line of colour glitches in a bar across the top a few millimetres wide.
This is getting worse. I hope that hasn't done some permanent damage

 

[Bee_Dee_3_Dee]

check the PM i just sent

 

[jamiewhite]

I'm not really sure where to start but I felt the need to come back and explain what's happened since. Unfortunately, I can't tell you the cause and my solution is hardly plausible.

As it is at the moment, I have been fully without my PC for two whole months and have endured either no PC or a broken one for 11 months now.
It's almost my 1 year anniversary... of the worst year of my life.
A lot of other crap has gone on this year besides this but as if it really needed to when a PC you've spent £1000+ goes belly up. I've moved back home with my parents because the stress and money has gotten a bit much. So I'm able to borrow their laptop for this thankfully.

I ended up RMA'ing my Gigabyte motherboard as their tech support had me flash the last official Award bios and that caused my PC to stop booting altogether. So without a PC I was very limited and have had to do everything since on my trusty Moto G, thank Google for small favours.

I have so far RMA'd my Corsair RAM and Gigabyte motherboard, the RAM was replaced like for like with a new set whereas the motherboard was returned to me "fixed". I'll give it to Gigabyte, it booted which was a bonus. But having waited a fortnight to get it back, I was disappointed that that was about all it did successfully still. The computer works seemingly fine via HDMI (it previously had the same issues as via the gpu) it boots and is usable, though installing the graphics card in any PCIe slot causes the PC to not boot. Gigabyte claim to have tested the PCIe slots and everything else regardless.

Anyway, they've been awesome enough to offer a like for like swap but with a fresh new motherboard. With Corsair having done the same with the ram, and hopefully I can convince them to do the same with the PSU, my plan currently is to sell all of the bits from my old PC and start afresh.
I know it's silly but this thing is cursed >_<

I've tried my best to learn everything I need to and more in my time loving PC's but I don't know all that much and am still learning, I don't have endless (or any) amounts of money or easily available spare parts. I've used my partners PSU to see if mine is the issue and everything stayed the same (i.e. still broken). I don't have a spare CPU or motherboard and can't afford to "go to a local professional", as overclockers who I originally purchased all this from have told me to do. I'm still in warranty for everything except the CPU mind you, but they don't care. By my count 6 of 8 things I purchased on that fateful day in 2012 have failed and all of them came from overclockers.

Right now, I'm just so stressed and anxious that I'm taking a break. I'm planning my new (re)build and will probably be posting and asking for some advice regarding that. But the best case scenario here is, that I get as many parts exchanged for new as possible to maximise my return to invest in the new PC and just bite the bullet.

Thanks for all your help guys, this was just way over my head and unfixable without having a tech guy with loads of spare gear on hand.

 

[junkeymonkey]

seemed like a few days ago a guy with the same board series could not get his new card to work on it [Z68 chipset] ??? maybe something up with that chipset and new cards or there uefi bios cant be updated to accept newer cards ???

http://www.tomshardware.com/answers/id-2364162/motherboard-detect-graphic-card.html