Is my graphics card faulty?

Hi all,

As an early Christmas present, my EVGA GTX 280 has seemingly decided to give up on me. I was playing a game of AOE 3 with a mate a couple of hours ago, when I got the following on the screen about 5-10 minutes into the game:

http://i210.photobucket.com/albums/b...1/IMG_6147.jpg

The 'white' box is where the mouse pointer was, and moved around accordingly.

We then gave Borderlands a try, and the same problem occurred, the first time in the menus, and the second time in-game after 2 minutes of gameplay. A shot of the second occurrence in Borderlands:

http://i210.photobucket.com/albums/b...1/IMG_6149.jpg

Note that the patterns are different, I'd guess that would be due to the 2 different applications.

The monitor is a 24" Samsung 245B.

After the issue in AOE 3, I cleaned off the drivers with driversweeper, installed them again, but it clearly did not help.

Does this look like a hardware issue? I'm guessing so...the card has just under a year of warranty left though :)

As an aside, over the last few weeks (since I installed Windows 7 in October), the card's fan has occasionally failed to throttle back after turning on the computer. There is then no picture on the screen, and upon a restart (where the card then 'starts' fine), I get an 'Overclocking Failed' warning on POST, which necessitates another reboot. Everything then works fine.

Another issue that happens very rarely, is that on boot into W7, very occasionally the animation is very jerky, for instance minimizing of windows/moving windows around etc, is not fluid like it would normally be.

The power on/jerky animation issues are rare, happening sometimes 2 days in a row, and sometimes not happening for a couple of weeks. I've talked to EVGA about these 2 problems, and they have no real explanation for it, other than that my CPU overclock might be causing issues. My specs are in my profile.

The first problem that I described is new, however, and has never happened before with this card.

My thanks to anyone who can be of any help!

mm, it could be a hardware issue, but I'd try it out with a few other applications first. Some other games, or benchmarks like 3DMark.

I take it when you reinstalled the drivers you used the latest ones off nVidia's site for Windows 7?

Quote:

---

Originally Posted by DevAnubis

mm, it could be a hardware issue, but I'd try it out with a few other applications first. Some other games, or benchmarks like 3DMark.

I take it when you reinstalled the drivers you used the latest ones off nVidia's site for Windows 7?

---

I'm running driver 182.50, as that is what works best for me with FSX, for which I built this PC. I know it's a little out of date, but i've had zero problems with it so far - all games that I play have worked flawlessly under this driver and windows 7.

Strangely, since 24th December, I've enjoyed a few hours of Borderlands with absolutely no trouble at all. There has been no recurrence of this problem over the last few days, and a few hours of ATITool/rthdribl/OCCT showed up nothing at all.

Could this just be a strange anomaly or a taste of things to come? :O_o1:

that really looks like memory overheat on your card to be honest.

Quote:

---

Originally Posted by Jay

that really looks like memory overheat on your card to be honest.

---

The card does run very hot, though that seems normal for the GTX 280. It's just weird that it's never done this before, and that it's run fine the last couple of days.

I'm just worried that if I send it off to EVGA, they probably won't be able to replicate the problem(s), and just send it back unchanged.

I turned on my PC a short while ago and got this more or less as soon as it reached the desktop:

http://i210.photobucket.com/albums/b...1/IMG_6156.jpg

The white box, unlike when it crashed during AOE 3, did NOT move when I moved the mouse.

I haven't had any problem (bar this) since 24th December, despite quite a lot of gaming, and 16 hours of continuous ATITool artifact scanning + rthdrible a couple of nights ago which showed up NOTHING.

The GPU Core temp in hwmonitor doesn't seem abnormally high for my particular card - it's running ATITool at the moment, at 80C. Idle temp is 47C.

Jay said that it looks to him like a memory overheat - that seems unlikely now that it happened immediately after a startup from cold.

What are the chances of a hardware defect even with no errors in ATITool/rthdribl?

Try running with the side-off your case to make sure there is maximum cold air getting in there and all the warm air getting out. Also you could risk taking your card apart and making sure all the RAM heatsinks are seated properly etc.

Quote:

---

Originally Posted by Terbinator

Try running with the side-off your case to make sure there is maximum cold air getting in there and all the warm air getting out. Also you could risk taking your card apart and making sure all the RAM heatsinks are seated properly etc.

---

My case is a HAF 932 with a 230mm intake fan on the side, and when I built the computer, I ran it with the side off for a couple of days, and I'm pretty sure the GPU actually runs cooler with the side on.

I'd rather send the card off to EVGA than dismantle it, it's a lot of money for me to be risking.

Does anyone think this looks like it could be a software issue?

Quote:

---

Originally Posted by collettnj

My case is a HAF 932 with a 230mm intake fan on the side, and when I built the computer, I ran it with the side off for a couple of days, and I'm pretty sure the GPU actually runs cooler with the side on.

I'd rather send the card off to EVGA than dismantle it, it's a lot of money for me to be risking.

Does anyone think this looks like it could be a software issue?

---

Have you tried the latest drivers?

Quote:

---

Originally Posted by Terbinator

Have you tried the latest drivers?

---

I use driver 182.50, because it is the driver that works best with FSX, which is my primary use for this PC. I've been running with this driver for 2 months, and have not had any problem like this until a few days ago.

I would try the latest one, but given that the PC just lasted 6 days without a single problem, and never experienced the problem for 2 months prior to that, it'd be a very lengthy process to test the newest ones! (ie, I could end up having to wait a long time for the issue to occur, during which time I would not be able to put the PC to its primary use!)

The first time this happened, on Christmas Eve, I wiped off the drivers with driversweeper, and downloaded and installed the 182.50 drivers again. The problem happened again a few minutes later when I was trying Borderlands.

Could this kind of corruption/visual be associated with a software problem?

does seem like an overheating issue to me, i had an 8800GT from evga that did that in the past. i changed the cooler flashed the firmware and it was much better.

by the way the cause of the problem for me was the fan started to get cranky and would go higher than 60% or something like that so then it kept overheating

I'd crack on with an RMA tbh.

Follow their instructions to the letter and send a couple of screen shots printed with it :)

Hi chaps, I emailed EVGA on Saturday - the problem has worsened, and I've pretty much narrowed it down to the graphics card I hope.

It seems that the issue occurs predominantly when the card is cold/heating up from cold - on Saturday I opened my window wide for a couple of hours in the morning, letting the room cool down to probably single figure temps, then turned on the PC. It booted fine (4GHz i7 at in the 20s on air :rolleyes: ), but the display corrupted as soon as I opened FSX.

I then set the bios to stock settings, removing the overclock, and let it cool for an hour. Again I booted it, and it crashed again as soon as I opened FSX. I hope that pretty much removes the CPU overclock from the equation.

I then tried an ATI 4550 in the PC, still with the room freezing cold, and there were no issues at all. The 4550 is of course in a completely different league to the GTX 280, but is the only other graphics card available to me at this time.

I emailed EVGA with plenty of photo evidence - I'm just hoping that they can replicate the issue!

Thanks for your help :)

An update on my situation here -

I sent the card off on the 13th, it arrived on the 18th, and they sent a replacement the same day, which arrived on 21st. Excellent service on their part.

The replacement, however (an identical card), has a problem of its own.

I turn on the PC, get a display on the monitor after 12 seconds (seems normal enough), but the graphics card fan doesn't fire up until 16 seconds after I press the power button. This seems a long time - my last card switched on pretty much instantly.

Once the fan spins up, it is stuck at 100%. It does not throttle back at all; even once W7 has loaded completely.

If I go into rivatuner, and open up the hardware monitoring window, I get the following:

http://i210.photobucket.com/albums/b...rivatuner1.png

If I then try to change the fan speed in rivatuner (in this case I've set it to 50%. I can hear that it has derated) the graphs do the following:

http://i210.photobucket.com/albums/b...rivatuner2.png

In the cases of both screenshots I've let it run for a couple of minutes to give a decent idea of the fluctuations in the values.

A few figures off the graphs:

Core temperature has a range of -119 to 108 degrees C

Ambient temperature is a similar range

Fan speed RPM is approx 200 (though varying by +/- 20) (the flat bit is ~3000)

Fan duty cycle varies between 39 and 85, though there is no change in the sound, it's making.

Sensor supply voltage varies between 1.73 and 3.47. The flat section is 3.26.


There is clearly something wrong here (duh). I've tried the card in 2 different PCI-E slots. No change. Sometimes the card shows up in HWMonitor, sometimes it doesn't.

I tried FSX, which ran very smoothly. No instablility there.

I'll give EVGA a ring on Monday about this, but does anyone have any suggestions as to what I could try in the meantime?