Gigabyte HD Radeon 7970 Glitches on Screen

Status
Not open for further replies.

SnoochieBooch

Honorable
Nov 24, 2012
15
0
10,510
Hello, I've recently upgraded to a Gigabyte HD Radeon 7970 stock OC'ed to 1000 MHz from an Nvidia GTX 560 ti and i continue to have these glitches pop up on screen either in the form of lines or checkered blocks much like this: http://i.imgur.com/yDtl25B.jpg

My current setup is

OPERATING SYSTEM: Windows 7 Home Premium 64 bit

CPU TYPE: AMD Phenom(tm) II X6 1090T Processor

CPU SPEED (GHZ): ~3.7

SYSTEM MEMORY (GB): 16

VIDEO CARD MODEL: AMD Radeon 7970

CATALYST VERSION: 13.2

DESKTOP RESOLUTION: 1920x1200

HARD DISK SIZE (GB): 1863.01

HARD DISK FREE SPACE (GB): 1373.88

POWER SUPPLY: Antec High Current Gamer 750 Watts

Any and all help will be much appreciated.
 
Solution
Yes, it's possible that it's a bad card. What I'd suggest first is opening up your registry and doing a search for nvidia or related terms to see if anything was left behind after the uninstall. You could also/instead use a program such as Revo uninstaller or CCleaner and there are a few other options too such as Driver Fusion, but if you don't want to install third party programs just to deal with drivers, running a few registry searches should do the job just as well.

If anything related to Nvidia graphics pops up in the search, then it probably needs to be deleted.

Still since you got the card so recently, it might be easier to just send it in for an RMA. If its already damaged, then no amount of software tricks will fix it and...
How drivers are replaced when you change cards can cause issues. For example, uninstalling drivers normally usually leaves traces of the uninstalled drivers around and that can cause issues. Do you know for sure that you did a proper driver replacement when you upgraded your cards?

Excessive heat is also a common cause for artifacts and other glitches. If just about any component of a graphics cards gets out of safe temps, which can happen (although it's usually caused by a faulty card or ambient temps being too high for the card to handle in a given situation), they can get damaged over time and start to fail. Unfortunately, except for the GPU temperature, most components on a graphics card do not have temperature sensors nor safe temperature specifications, so it can be difficult to troubleshoot heat issues.

One way is to significantly underclock some parts such as the GPU and/or the memory. Underclockign the memory should bring memory temps down and underclocking the GPU and memory should bring temps on some other components such as the VRM down. Checking to see if the glitches still appear after underclocking (potentially also or replaced with undervolting) can help narrow down the suspect components. If acted on quickly enough, assuming that excessive heat is the issue, it can be possible to fix the issue before any severe damage has been done.

Other issues are also possible. For example, some cards simply fail for little apparent reason (this is rare, but does happen). If the card is permanently damaged, then an RMA may be necessary.

If you could give further cooling information about your case, that would be helpful.

Another less likely but possible scenario is that your PSU is failing. How long have you had it?
 

SnoochieBooch

Honorable
Nov 24, 2012
15
0
10,510
Would installing the new drivers before uninstalling the nvidia drivers maybe have caused problems? Also even under heavy load the GPU temperature has never gone over 53 degrees celcius and usually idles at about 40 degrees.
 
Yes, that could have caused your issues.

If it really never goes over around 53C, then GPU temp shouldn't be an issue. However, I suggest making sure that you check the GPU temperature with at least three different programs. Sometimes, some program will report it differently from the rest. IDK why they don't all agree on temps, but I find that it's usually best to get multiple *opinions* on what the temps are just to be sure.
 

SnoochieBooch

Honorable
Nov 24, 2012
15
0
10,510
So I've gone through and fully uninstalled all radeon drivers and did a full reinstall with the same effect. I also have used CCC and MSI Afterburner to do temperature checks. Is it just possibly a bad card? I've only had it for about 3 days so it's not an issue to return it.
 
Yes, it's possible that it's a bad card. What I'd suggest first is opening up your registry and doing a search for nvidia or related terms to see if anything was left behind after the uninstall. You could also/instead use a program such as Revo uninstaller or CCleaner and there are a few other options too such as Driver Fusion, but if you don't want to install third party programs just to deal with drivers, running a few registry searches should do the job just as well.

If anything related to Nvidia graphics pops up in the search, then it probably needs to be deleted.

Still since you got the card so recently, it might be easier to just send it in for an RMA. If its already damaged, then no amount of software tricks will fix it and it's seeming more and more likely that it is faulty.
 
Solution

SnoochieBooch

Honorable
Nov 24, 2012
15
0
10,510
Just an update on the card situation. I went through and made sure everything was beloved patriot and span in the registry, and still to no avail. So i decided to exchange the card for a new one and I've had no issues. Seems it really was just a faulty card. Thank you for your time and replies though.
 
Status
Not open for further replies.