Is my computer killing video cards???

[Derek12]

I was playing chess titans and suddenly got this mess



Even Windows Aero artifacts I cleaned drivers but the problem remains so I think it's the video card
This is my second 5450 that seems failing
is this bad luck, my computer killing them or...????

Many thanks

[qubit]

Might be the PSU, but this seems unlikely, even though it doesn't seem to be one of the big brands.

I think you've just got unlucky with video cards most likely.

[de.das.dude]

check your RAM slots. clean them with a brush.

[Derek12]

Originally Posted by qubit

Might be the PSU, but this seems unlikely, even though it doesn't seem to be one of the big brands.

I think you've just got unlucky with video cards most likely.

Many thanks, then I will return it and get, maybe another 5450, or a 6450 or a GTX 520 I hope they last longer lol

Originally Posted by de.das.dude

check your RAM slots. clean them with a brush.

I tried both slots (I don't have another stick but I memtest'd it and was clear) but the problem remain. I removed the 5450 and now using the IGP and seems fine for now

Oh I forgotten the problem is increasingly worse now my computer got that





Many thanks!

[MRCL]

Mother of God.

Um yeah your video card seems to not like you at all. Can you try it in another computer, maybe a friends, to see if it acts the same there? If yes, its your card. if no, the problem lies soemwhere else.

[qubit]

Man, that second screenie is really bad.

Forgot to say: do you have another video card to check against? Also, you might like to try that card in another PC, for extra certainty.

[JrRacinFan]

D4 to a2 Just playing. I'd check the card in another's PC before calling it bad. ALTHOUGH HIGHLY UNLIKELY There may be an issue with the PCIe slot itself.

[trickson]

Looks to me as your video card is about to die if not dead already .

[Suhidu]

I had a laptop(Mobility Radeon 9000 64MB) that a few years back developed similar looking issues, a little at first and eventually a screen-wide pink-ish mess! (I think underclocking may have alleviated it at first.) It even corrupted the BIOS screens. Anyway, it wasn't the System RAM nor anything else user-replaceable, and so after Googling allot about video card artifacts by the GPU or corruption in bad video memory*, I assumed it was the dedicated video card's memory going. This video card was part of the motherboard itself and the circa-2003 product was no longer under warranty, so I haven't used it sense.
*http://www.google.com/search?q=video...ient=firefox-a
http://www.playtool.com/pages/artifacts/artifacts.html

That's odd in your case, two 5450s going like that. Were they the same model? Does that model get bad reviews? An insufficient cooler for example, could cause damage to a GPU or Video Memory.

[robal]

What you see here is failing GPU memory.
Nothing can be done to fix that.

Temporary solution (available on some video cards) is to underclock GPU mem.
But the problem will return in future anyway.

I don't think your PC (PSU) kills those cards.
Most probably you're just unlucky.

Cheers,

[newtekie1]

If the cards are passive, maybe there isn't enough airflow in the case and they are overheating, slowly dying over time?

[Derek12]

Many thanks to all
I will reply one for one hehe

@MRCL, I don't have another computer there, the rest of them are all laptops but maybe I can get a neighbour to let me try it on his computer! I will post the results when the tests are done unfortunately he uses a IGP so I can't check another vid card on mine computer

@qubit and JrRacinFan: unfortunately I don't have another video card (I will ask some friends/neighbours anyway), but I tried the IGP and all is fine, must I assume the m/b (minus the PCIe port which I will try to check) is fine too? (I hope)

@Suhidu that would suck I had a Laptop w/dedicated video that also failed, wish they made removable VRAM for laptops (or removable video cards like the MXM ones!

Yeah, they are the same model, PowerColor HD5450 full height video card with 512 MB. I didn't saw many reviews of them but seems to be positive
According to the article and the first screenshot indeed is a video memory issue (unless it's the motherboard/PCIe port)

@robal Good idea!, I will try to install MSI afterburner** and underclock it and will post the results that also would pinpoint the video card or motherboard as the culprit here.

@newtekie effectively they are passive, though the temps seems to be fine (60º load, 35º idle) I tested the case exhaust fan is working

___________

**I said "try" because the problem increases alarmingly: now I am getting the "Driver stopped responding and has recovered blah blah blah
And also now the PC hangs randomly showing a black screen with thin blue vertical lines, and then it reboots and a message appears saying "Windows has recovered from an unexpected shutdown" I assume that this screen of vertical lines was indeed a BSOD

[qubit]

You can't completely assume the mobo is fine, but corruption problems like this are 99% of the time the graphics card. If the PCI-E slot was corrupting data, the whole PC would be likely to lock up.

[Derek12]

Originally Posted by qubit

You can't completely assume the mobo is fine, but corruption problems like this are 99% of the time the graphics card. If the PCI-E slot was corrupting data, the whole PC would be likely to lock up.

Update: I managed to underclock the video card RAM from 400 to 200 and the artifacts were gone thought Atitool still detect them, but the desktop is now clear, they start at 350 lol, should I assume definitely is a video card memory issue and not motherboard PCIe, right?
if it is the case I will return it


Many thanks

[d3fct]

i would say yes, corrupted memory. just from the pink screen shot, I've heard it be reffered to as the pink screen of death. I have had 1 or 2 nvidia cards take a shit on me just out of the blue, and ended up be corrupted memory. it does happen, I would look into replacing it.

[qubit]

Originally Posted by Derek12

Update: I managed to underclock the video card RAM from 400 to 200 and the artifacts were gone thought Atitool still detect them, but the desktop is now clear, they start at 350 lol, should I assume definitely is a video card memory issue and not motherboard PCIe, right?
if it is the case I will return it


Many thanks

Definitely return that card. That's a pretty definitive test you ran there.

[DigitalUK]

did you buy the cards from the same place, maybe they have a bad batch or something. ive used loads of the HD5450 passive cooled for HTPC's and basic desktops (in slimish cases) never had one do that, i use a 6450 passive (since release day) in my own htpc never had any problems and that thing is left on 24/7 most days.

[Derek12]

Many thanks to all, I returned it and got a GTX 520 which will arrive in Wednesday-Friday, meanwhile, I am using the IGP
@DigitalUK, Correct, I bought them at the same site, maybe you are right, that or very bad luck

[Live OR Die]

Looks like your GPU memory is failing i've seen this problem on my old 8800 Ultra few days after no video at all just a black screen. Well you know if your new nvidia card ends up doing the same thing there a problem with your board.

[Derek12]

Originally Posted by Live OR Die

Looks like your GPU memory is failing i've seen this problem on my old 8800 Ultra few days after no video at all just a black screen. Well you know if your new nvidia card ends up doing the same thing there a problem with your board.

Many thanks, lets see how the nvidia one works
Shortly before returning it, the BIOS posts was also affect, showing scrambled texts like "virijiynk" instead of verifying [DMI pool data] so, imagine!

[Derek12]

UPDATE:

I am beginning to think my motherboard is on its way out and is really killing video cards, now I suddenly got this while browsing here with my GT 520, but seems to be temporary, lasted a few seconds, and now is fine again, but I will try to game and see if artifacts there:



Glitch on Firefox 10 alpha? I was testing the new 10 alpha (unlikely because the corruption was even in the taskbar) I will use Opera and see if happens there
Motherboard bad?
Bad PSU? unlikely because the voltages seem to be correct!

I am pretty sure 3 cards, two of them are different vendor, can't be bad

[DRDNA]

I would accuse the PSU now...how long did it run b4 the corruption?
If that PSU has a terrible ripple it could do something like that.

[Jegergrim]

Indeed, I think we can safely assume that by now the cards themselves are fine, if you get a chance try to test the PCI-E port with another card, perhaps a cheap old card which you overclock slighty. If the problem doesn't occur, you should definately accuse the PSU, and moreover you should STOP using the PC till you replace the PSU, as you don't want to damage more components than nescessary

[JrRacinFan]

I mentioned previously in the effect to RMA that motherboard. Now it's clear you may need to find a different case and power supply as well.

[AsRock]

Originally Posted by Derek12

Many thanks, then I will return it and get, maybe another 5450, or a 6450 or a GTX 520 I hope they last longer lol



I tried both slots (I don't have another stick but I memtest'd it and was clear) but the problem remain. I removed the 5450 and now using the IGP and seems fine for now

Oh I forgotten the problem is increasingly worse now my computer got that

http://img.techpowerup.org/110918/Capture027.jpg



Many thanks!

Last time i seen that was with my 2900XT when the ram chips failed. Make your own conclusion from that.. But i would say RMA that video card.

Although as you said you think it might be the mobo i suppose it could be you got any spare parts to test ?.