Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Sign in to follow this  
Danarchy

Vidya Issues

Recommended Posts

So I'm having issues with my video card, and I'm about at the end of my rope with it now. I have a GTX 460 and I'm getting all sorts of weird video issues when I play games. Extra planes branching off into infinity, random colors on certain models, missing models, all kinds of stuff. It seems to be random in each game I play just what's wrong, too.

It seems to be a driver issues, but I keep updating and reinstalling them and it doesn't seem to fix it. Each time, I use Driver Sweeper to get rid of the old drivers, then reinstall it with a clean install (probably redundant). I keep searching online for help, but it's mostly useless information and people calling people noobs for not being able to troubleshoot their own card issues. I did once find some guide some guy wrote for doing a complete clean reinstall of drivers properly, and I followed it. It worked...for about a week, then the issues came back. I've even tried installing different versions of the driver, both old ones and beta versions, and for some reason i get completely different issues with each version I install.

So, I ask you people of Doomworld, am I fucked here, or what? Do I need to get a new card (or hope I can a replacement), is my card incompatible with my computer, or are these just silly driver issues that I'm too much of a noob to figure out? Help me, fellow Doomers, you're my only hope!

Share this post


Link to post

You're quite possibly fucked. That kind of artifacting is often dying hardware just messing with your head. Sometimes it's just overheating hardware, so check for that too.

Share this post


Link to post

Second what Aliotroph said.

My anecdote is when I got a new computer with a geforce 8800gt in 2007. Up until may '11 it worked perfectly until it started artifacting then crashing 30min into games. Underclocking/changing drivers didn't help. Then the symptoms happened sooner through the next few days, until the second it boots into XP/7 it crashes with a garbled desktop.
Replaced with radeon 6870 that works fine, so my 8800gt is crapped out.

Share this post


Link to post

Hmm, that sucks. I've only had it about 6 months.

It might be possible that it's overheating. Maybe it's because I keep my computer running all the time?

Share this post


Link to post

Did you update the drivers at some point before the issues began? I had some nasty shit going on with my 8600 GT with the few latest nVidia drivers, and scrounging the nVidia forums it turned out to be a widely acknowledged problem. Going back to older drivers fixed everything.

So the very first thing I'd do is check the nVidia forums and make sure whether the driver version you're using is reported to have problems with your card.

Share this post


Link to post

A good point. Occasionally they break drivers badly enough that it's worthwhile to use older ones.

If that doesn't work then you can check it for temperature. The drivers usually have their own tool for that and there are often third-party ones that will work. If it's too hot (might need to look up what constitutes too hot) then see if the heatsink on the card is full of dust. They clog easily if a lot of dust gets into your machine. It's easy to blow out the dust. Easier if you have canned air. Normally running a computer continuously shouldn't cause any part of it to overheat.

Share this post


Link to post

The rudimentary diagnosis that hasn't been mentioned yet: is the fan still working? But like Aliotroph said it could be dust and shit clogging up the heatsink. Being summer and presumably warm/hot where you live won't help anything. Whatever the cause it definitely sounds like an overheating GPU to me though.

EDIT: If it's a dead fan and your card isn't under warranty then you could contact the vendor for a replacement fan. Cheaper then buying a whole new card, natch. Or I guess you could ghetto some other fan on there if that's your thing.

Share this post


Link to post

The GTX 4xx series Nvidia Fermi's ran really hot. So having completely ruled out drivers as suggested by Jodwin, the prognosis isn't very good.

Download GPU-Z: http://www.techpowerup.com/gpuz/

Verify your GPU fan is working.

Check your temps and fan speed @ idle, and then boot up a game or something and see where it's at when it's load. Alternatively, get Furmark. Run the stress test and put your GPU @ 100% load and see how toasty things get.

Verifying the fans works, clean the GPU heatsink as suggested. You can also get a program like Riva Tuner to manually adjust your cards fan speed. Might turn it into a noisy jet engine. But if summer is doing a number on your ambient room temperature, it might just be what's doing a number on your card.

Failing that, you'll probably need a new card. Unless you can get rid of the artifacts by under clocking as a final last ditch resort. Also check the overall temperature of your other components, EDIT: with AMD Overdrive. You got AMD right? might need more airflow/fan(s) in your case if permissible.

Share this post


Link to post

Also one thing to try if you think its overheating is to take out the case door and have it near a space that well-ventilated or even near a small floor fan so that the case parts cycles the heat sufficiently.

Share this post


Link to post

One more for cleaning out the HSF, but I'd recommend using MSI Afterburner instead of Rivatuner since RT hasn't been updated in a long time (mostly because the guy updating RT moved on to Afterburner).

What kind of PSU do you have? If it's an older PSU, maybe it doesn't have enough amps.

Share this post


Link to post
POTGIESSER said:

You can also get a program like Riva Tuner to manually adjust your cards fan speed. Might turn it into a noisy jet engine. But if summer is doing a number on your ambient room temperature, it might just be what's doing a number on your card.

Seconded. When I first got my 275 it was overheating because the fan speed was stuck at 40%. Installed Riva Tuner and set it to up the speed based on the temperature. Everything was fine after that, except for the noise when it hits 60%, of course.

Share this post


Link to post
POTGIESSER said:

The GTX 4xx series Nvidia Fermi's ran really hot. So having completely ruled out drivers as suggested by Jodwin, the prognosis isn't very good.

Download GPU-Z: http://www.techpowerup.com/gpuz/

Verify your GPU fan is working.

Check your temps and fan speed @ idle, and then boot up a game or something and see where it's at when it's load. Alternatively, get Furmark. Run the stress test and put your GPU @ 100% load and see how toasty things get.

Verifying the fans works, clean the GPU heatsink as suggested. You can also get a program like Riva Tuner to manually adjust your cards fan speed. Might turn it into a noisy jet engine. But if summer is doing a number on your ambient room temperature, it might just be what's doing a number on your card.

Failing that, you'll probably need a new card. Unless you can get rid of the artifacts by under clocking as a final last ditch resort. Also check the overall temperature of your other components, EDIT: with AMD Overdrive. You got AMD right? might need more airflow/fan(s) in your case if permissible.

Ooh, thanks. Looks like my fan is working okay at the moment. I'll check it later with games running and see what happens.

My PSU should be fine. I remember checking that before I bought the card.

I was checking out the NVidia forums and did see that my card series has a lot of issues. It seems everyone else (of the dozens of threads I've read) was having different problems than me, though (or were just getting trolled instead of helped).

It's weird because my card was running great for a while, then it started getting glitchy and artifacty, so I updated my drivers. Then it was working fine for another couple of months, then it got glitchy again, and soon after NVidia released updated drivers, and that worked for another month...and that's where my current problems start. Something tells me it might be a driver issue and I'm doing something wrong when I'm reinstalling them.

Share this post


Link to post
Danarchy said:

It's weird because my card was running great for a while, then it started getting glitchy and artifacty, so I updated my drivers. Then it was working fine for another couple of months, then it got glitchy again, and soon after NVidia released updated drivers, and that worked for another month...and that's where my current problems start.

Either your video card's dying a slow death or Nvidia are trying to stay on top of a hardware issue that refuses to be tamed - or both.

Share this post


Link to post
Danarchy said:

Something tells me it might be a driver issue and I'm doing something wrong when I'm reinstalling them.

It's always been good practice to completely uninstall your existing video drivers and restart when prompted before you touch any new ones. You shouldn't have to take any further precautions, though.

Share this post


Link to post
DoomUK said:

It's always been good practice to completely uninstall your existing video drivers and restart when prompted before you touch any new ones. You shouldn't have to take any further precautions, though.

That's what I have been doing. Guess it's not drivers, then.

Share this post


Link to post

Who's the manufacturer, out of interest? Sounds like one to avoid in the future for such a new card to show signs of dying, if cooling, PSU and driver issues have been completely ruled out. Considering the plethora of nvidia manufacturers out there it's not like you only have the option of buying the same card from the same company.

Share this post


Link to post

Tried underclocking yet? Should try it for laughs.

Especially if it's a "factory overclock" card. Consider dropping the clocks to specs of a non OC. Could be a card that barely passed the factory OC binning process, and now it's acting up.

Share this post


Link to post

I once got a Gainward Bill PCIe 7600GT card to die a sudden and very violent death. At first, I simply walked into the room and found my box had shut down, and trying to turn it on had no result (the fans did half a spin, then died). I started removing hardware, and it rebooted as soon as I took out the video card...not a good sign.

I tried it in the mobo of a friend of mine, which DIDN'T refuse to power up, and made us realize how dangerous a stay-closed failure can be: as soon as we hit the power button, we were greeted by a shower of sparks and smoke, coming from a defective power transistor that had started to glow red-hot.

Luckily, my friend's mobo survived, but later I discovered that mine had a permanently damaged PCIe slot (and also its AGR slot) which mean that I could not use either PCIe or AGP cards for long: as soon as I powered up ANYTHING using OpenGL or Direct3D, they locked the system down.

Only vanilla desktop use worked, software-rendered games ;-) and/or PCI-only cards.

Share this post


Link to post
DoomUK said:

Who's the manufacturer, out of interest? Sounds like one to avoid in the future for such a new card to show signs of dying, if cooling, PSU and driver issues have been completely ruled out. Considering the plethora of nvidia manufacturers out there it's not like you only have the option of buying the same card from the same company.

Wait, not all NVidia cards are made by the same manufacturer? How do I tell who made it?

POTGIESSER said:

Tried underclocking yet? Should try it for laughs.

Especially if it's a "factory overclock" card. Consider dropping the clocks to specs of a non OC. Could be a card that barely passed the factory OC binning process, and now it's acting up.

Hmm...I forget how to under/overclock. Is it on the hardware? I think I saw a switch somewhere.

Anyway, I ran that program and it told me the card was running in the mid 50s Celsius, which is apparently optimal temperature. So I guess it isn't overheating.

Share this post


Link to post

Use Riva Turner or this to play with the clocks:

http://downloads.guru3d.com/MSI-AfterBurner-download-2392.html

After burner would probably be better choice.


mid 50's celcius? Sounds reasonable for idle temps. How about at Load? Have you tested to see how high the temperature is at load?

Try Firing up a game or benchmark and see how hot it gets.

Danarchy said:

Wait, not all NVidia cards are made by the same manufacturer? How do I tell who made it?


You should be able to gather that info from GPU-Z or after burner. Even display adapter should tell you. They're all NVIDIA cards but different companies sell their cards with their brand. There's reference spec cards from NVidia, but these companies will also make their own models that deviate from reference spec. Should probably say something MSI, Evga, Asus, Zotac, unless it's NVIDIA OEM in a prebuilt computer like HP or something.

Share this post


Link to post
Danarchy said:

Wait, not all NVidia cards are made by the same manufacturer? How do I tell who made it?

If there's no other indicator about it, the card itself should have a sticker or two with the manufacturer's name, or it might be printed somewhere on the card. Or if you have the package or any of its contents stored somewhere, etc...

The manufacturer can be pretty important, mostly because NVidia/Ati only provide the actual processing components, all the other stuff such as power connectors, fans, etc. are left to the manufacturer. Some manufacturers use shitty components that tend to fail while some others use quality ones.

Share this post


Link to post
POTGIESSER said:

Use Riva Turner or this to play with the clocks:

http://downloads.guru3d.com/MSI-AfterBurner-download-2392.html

After burner would probably be better choice.

Cool, but I don't have any idea what I'm doing here. It's set to the middle right now. Should I just turn it down and underclock it then?

mid 50's celcius? Sounds reasonable for idle temps. How about at Load? Have you tested to see how high the temperature is at load?

Try Firing up a game or benchmark and see how hot it gets.

Mid-high 50s was at load. It's in the upper 30s when idle. It's not too cold is it? :P

You should be able to gather that info from GPU-Z or after burner. Even display adapter should tell you. They're all NVIDIA cards but different companies sell their cards with their brand. There's reference spec cards from NVidia, but these companies will also make their own models that deviate from reference spec. Should probably say something MSI, Evga, Asus, Zotac, unless it's NVIDIA OEM in a prebuilt computer like HP or something.

Okay, looks like EVGA. How do they rate?

Share this post


Link to post
Danarchy said:

Cool, but I don't have any idea what I'm doing here. It's set to the middle right now. Should I just turn it down and underclock it then?

Move the GPU clock sliders to the left in small increments (like, 5-10MHz at a time or something). Making massive adjustments could cause unwanted instabilities (maybe that only happens when you overclock, but best to be safe). Test & repeat until the chip is set to whatever the non-overclocked version runs at or when it appears to be stable. Might want to try lowering the memory clocks in the same manner if all else fails.

Danarchy said:

Okay, looks like EVGA. How do they rate?

EVGA has always been a reputable manufacturer. Looks like you could be the guy with the one-in-a-hundred-million faulty card :/

EDIT: Determining the card's BIOS rev. and flashing it to a newer version if there's one available might also be something you might want to try. I'd have to let someone else step in and instruct you on how to do it though. I did it once with a Geforce 7 I used to own, but I wouldn't trust my memory on that.

Share this post


Link to post

Unfortunately it sounds like the card is damaged so don't expect it to work for long, even underclocked. Eventually it will just die. For the record, my two EVGA cards are doing swell, both from 2004.

Share this post


Link to post

Oh hey, another person with GTX 460 issues. In my case, I don't get display glitches ingame, but with some games (Crysis 2, but also some Source games) the display driver randomly crashes which often locks up the whole computer with artificats and looping sounds etc forcing me to reset. This happens regardless of driver version or OS. It's a Colorful GTX 460 1GB. I expected the card to be faulty right away - for it's only a couple of weeks old - but the manufacturer's tech support kind of treated me like an idiot and suggested things like reinstalling the OS and testing the card on a different mainboard - as if I had different hardware lying around everywhere.

What indeed does help is lowering the clock speed, as people here suggested as well. But I still have no idea what could be the cause of this issue. Some hardware incompatibility, mainboard? Or inconsistent PSU rates?

The card's performance is also quite ambiguous. Some games, like Crysis 2, run fine on max settings (minus the crashes), others, like TF2 for instance, have fairly poor fps regardless of the settings. And as you said yourself, the GTX 460 seems to cause problems for a lot of people.

I've even considered getting rid of the bastard, selling it to some unsuspecting victim on ebay. :(

Share this post


Link to post
neg!ke said:

I've even considered getting rid of the bastard, selling it to some unsuspecting victim on ebay. :(


Please don't do what I think you mean to do. A friend was going to sell his broken old Xbox360 on eBay in the same manner and I'll say to you what I said to him. It's thanks to people selling faulty stuff on eBay et al as working stuff that buying second hand hardware is a risky business in the first place. Advertise it as non-working, specifically outlining what appears to be wrong with it. Advertising it as fully-functional makes you a colossal douche.

Share this post


Link to post

Yeah, I know... But what if the card is okay and the problem is only compatiblity-related. I guess I could/should have it tested in a computer shop, but that would cost at least 15 bucks which would increase my potential loss even more. And seeing how many people encounter similar issues, there must be something to it?

Share this post


Link to post
Use3D said:

Unfortunately it sounds like the card is damaged so don't expect it to work for long, even underclocked. Eventually it will just die. For the record, my two EVGA cards are doing swell, both from 2004.

Oh. :(

Yeah, I heard there are issues with the 460 cards. I guess I made a bad purchase for my first video card.

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  
×