AnimeSuki Forums

-KarumA- · 2013-09-12, 16:29

Hello guys,

Last July I bought a GTX 680, the 4gb version by EVGA.
It was to replace an older card of mine which was getting out of date.

Anyway things went great, I was able to game properly. I am not overclocking it or anything and temperatures are always fine. Drivers downloaded fromt he website, not the beta ones but the latest official ones.

The fine part was till about a month ago. I was looking something up online, not gaming or anything, and my monitors flickered with in the corner the message that my display drivers stopped responding and recovered. I had to reboot, saw nothing faulty in my logs, had no BSOD anything. I figured hey, it can happen you know!

But since then this has happened every week. Not when I am gaming or stressing it out but at a random moment when I am about to press send for an email or when I open another program. Nothing that stresses the card out. Which is the problem; it games just fine, I ran furmark for a while and it passed. No this happens when I do nothing special at all and it annoys me.

Now what I tried; I tried looking for other cases and by God there are so many with all different causes that it doesn't help me one bit.
Some suggested use older drivers; so I uninstalled it clean and reinstalled older drivers, I was happy till about 7 days later when it suddenly happened again. So I decided to try the newest beta drivers because others had mentioned it, but today again it happened little more than a week later.

Here are also my other specs; 750w power supply which is enough, 8gb ram, AMD Phenom II x6, Asusteck mobo m4A87TF/usb3.

The GTX 680 uses even less power than my previous card which was I believe a GTX 280 or something.

This had been the first time that I ever bought a brand new gpu and frankly I don't know what to do about the problem.
I read up on the TDR errors on their website but the causes vary from a bad card to voltages to drivers causing it and so on (such as stated here; https://forums.geforce.com/default/t...dmkm-stopped-/)

But I don't know what to link to my specific problem. Whether it is because of a bad card; if so then how can it function when I am stressing it out. Or anything else.
Should I RMA it for a new one? Is there no reason to do so because perhaps it might not help (in case drivers cause the problem)?

Can anyone give me some advice on this?

Will add that this weekend I am going to try memtest to check the vram.
And I am keeping GPU-Z running now and keeping a log on voltages, temps etc. just in case it rehappens so I can take a look at that.

-KarumA- · 2013-09-13, 19:02

Stupid question now though; I noticed while checking the logs that every once in a while when not doing anything the gpu core clock would suddenly be maxed out and within a couple seconds lower again. Can this be a bad thing? I am figuring this because perhaps if it is loading too much at once it could cause a TDR.

sa547 · 2013-09-14, 07:07

Knowing your warranty period, you could RMA it if you didn't overclock the card.

Also, I'm not sure (since I'm now using a Radeon and Catalyst so I'm no longer keen about nVidia's products) but is the problem with the latest Forceware drivers (each update has a list of fixes they made) have been resolved?

-KarumA- · 2013-09-14, 08:51

Quote:

Originally Posted by sa547

Knowing your warranty period, you could RMA it if you didn't overclock the card.

Also, I'm not sure (since I'm now using a Radeon and Catalyst so I'm no longer keen about nVidia's products) but is the problem with the latest Forceware drivers (each update has a list of fixes they made) have been resolved?

The problem is if it is caused by drivers and I would RMA the card the same problems would just reoccur when I put in the new card. Which is what I am trying to figure out and having a hard time at since I know little about the possible causes and what conclusion to make of it. This because I read on other places the GPU memory clock thing being maxed is alright so I don't know..

As for the drivers, it depends what problems. Some cards have problems and some do not, for some people new released ones fix a lot of issues while for others it makes things worse.

Tiberium Wolf · 2013-09-14, 18:32

I think the best solution is you make a clean install of the OS and the drivers and see if that happens.

Oh btw, don't use beta drivers... just use the last stable version

-KarumA- · 2013-09-19, 01:59

I think my problem has been resolved. Last time I uninstalled it I ran CCleaner afterward before installing a same older driver version that had been giving me problems in the past but had been reported stable. But since then I haven't had any TDR crashes and it has been about a week now.

SaintessHeart · 2013-09-23, 10:39

Try using the 327.34 drivers. The ones between 314.xx and 32x.xx series have craploads of problem.

If not, try rolling back to the 314.xx series of drivers until Nvidia fixes their stuff.

Jinto · 2013-09-25, 03:05

Actually the problem is the myriad of cards using similar but not equal designs. NVidia provides the GPU and certain parts of the card, but can only recommend how OEMs design and assemble their cards.

Of course NVidia will build their drivers according to the reference design. Some OEM card manufacturers try to go cheaper in the design, which can then cause all sorts of trouble. That is why I typically buy cards from brands I trust to be equal or better than the reference design (e.g. ASUS or Gigabyte).

2013-09-12, 16:29	Link #1
-KarumA- (｡☉౪ ⊙｡) Author Join Date: Jul 2004 Location: In Maya world, where all is 3D and everything crashes Age: 36	GPU problems eventhough it is brand new Hello guys, Last July I bought a GTX 680, the 4gb version by EVGA. It was to replace an older card of mine which was getting out of date. Anyway things went great, I was able to game properly. I am not overclocking it or anything and temperatures are always fine. Drivers downloaded fromt he website, not the beta ones but the latest official ones. The fine part was till about a month ago. I was looking something up online, not gaming or anything, and my monitors flickered with in the corner the message that my display drivers stopped responding and recovered. I had to reboot, saw nothing faulty in my logs, had no BSOD anything. I figured hey, it can happen you know! But since then this has happened every week. Not when I am gaming or stressing it out but at a random moment when I am about to press send for an email or when I open another program. Nothing that stresses the card out. Which is the problem; it games just fine, I ran furmark for a while and it passed. No this happens when I do nothing special at all and it annoys me. Now what I tried; I tried looking for other cases and by God there are so many with all different causes that it doesn't help me one bit. Some suggested use older drivers; so I uninstalled it clean and reinstalled older drivers, I was happy till about 7 days later when it suddenly happened again. So I decided to try the newest beta drivers because others had mentioned it, but today again it happened little more than a week later. Here are also my other specs; 750w power supply which is enough, 8gb ram, AMD Phenom II x6, Asusteck mobo m4A87TF/usb3. The GTX 680 uses even less power than my previous card which was I believe a GTX 280 or something. This had been the first time that I ever bought a brand new gpu and frankly I don't know what to do about the problem. I read up on the TDR errors on their website but the causes vary from a bad card to voltages to drivers causing it and so on (such as stated here; https://forums.geforce.com/default/t...dmkm-stopped-/) But I don't know what to link to my specific problem. Whether it is because of a bad card; if so then how can it function when I am stressing it out. Or anything else. Should I RMA it for a new one? Is there no reason to do so because perhaps it might not help (in case drivers cause the problem)? Can anyone give me some advice on this? Will add that this weekend I am going to try memtest to check the vram. And I am keeping GPU-Z running now and keeping a log on voltages, temps etc. just in case it rehappens so I can take a look at that.

2013-09-14, 07:07	Link #3
sa547 Senior Member Author Join Date: Oct 2007 Location: Philippines Age: 47	Knowing your warranty period, you could RMA it if you didn't overclock the card. Also, I'm not sure (since I'm now using a Radeon and Catalyst so I'm no longer keen about nVidia's products) but is the problem with the latest Forceware drivers (each update has a list of fixes they made) have been resolved? __________________

2013-09-14, 18:32	Link #5
Tiberium Wolf Senior Member Join Date: Dec 2004 Location: Portugal Age: 44	I think the best solution is you make a clean install of the OS and the drivers and see if that happens. Oh btw, don't use beta drivers... just use the last stable version __________________

2013-09-23, 10:39	Link #7
SaintessHeart NYAAAAHAAANNNNN~ Join Date: Nov 2007 Age: 35	Try using the 327.34 drivers. The ones between 314.xx and 32x.xx series have craploads of problem. If not, try rolling back to the 314.xx series of drivers until Nvidia fixes their stuff. __________________ When three puppygirls named after pastries are on top of each other, it is called Eclair a'la menthe et Biscotti aux fraises avec beaucoup de Ricotta sur le dessus. Most of all, you have to be disciplined and you have to save, even if you hate our current financial system. Because if you don't save, then you're guaranteed to end up with nothing.

2013-09-25, 03:05	Link #8
Jinto Asuki-tan Kairin ↓ Join Date: Feb 2004 Location: Fürth (GER) Age: 43	Actually the problem is the myriad of cards using similar but not equal designs. NVidia provides the GPU and certain parts of the card, but can only recommend how OEMs design and assemble their cards. Of course NVidia will build their drivers according to the reference design. Some OEM card manufacturers try to go cheaper in the design, which can then cause all sorts of trouble. That is why I typically buy cards from brands I trust to be equal or better than the reference design (e.g. ASUS or Gigabyte). __________________ Folding@Home, Team Animesuki

2013-09-13, 19:02	Link #2
-KarumA- (｡☉౪ ⊙｡) Author Join Date: Jul 2004 Location: In Maya world, where all is 3D and everything crashes Age: 36	Stupid question now though; I noticed while checking the logs that every once in a while when not doing anything the gpu core clock would suddenly be maxed out and within a couple seconds lower again. Can this be a bad thing? I am figuring this because perhaps if it is loading too much at once it could cause a TDR.

2013-09-19, 01:59	Link #6
-KarumA- (｡☉౪ ⊙｡) Author Join Date: Jul 2004 Location: In Maya world, where all is 3D and everything crashes Age: 36	I think my problem has been resolved. Last time I uninstalled it I ran CCleaner afterward before installing a same older driver version that had been giving me problems in the past but had been reported stable. But since then I haven't had any TDR crashes and it has been about a week now.