Sign in with
Sign up | Sign in
Your question

Graphics anomalies slowly getting worse, please advise.

Last response: in Systems
Share
January 15, 2008 9:04:09 AM

For a while now I've been getting strange little errors at first i thought it was the game or a corrupt model file. I wasn't having any other trouble with my PC so the thought it was my own system didn't really stick.

However lately things have started going down hill quite fast and I'm pretty sure "something" will die soon. That's what I want help with really. I know something need replacing but there isn't a clear winner.

This is pretty much what I'm now getting quite regularly. Corrupted textures and polygons getting stretched in random directions, often by more than one vertex...

(couldn't find the syntax to shrink the image and I formatted recently and don't have any image manipulation tools back on)

Anyway, things didn't start out this severe, it was often just a very infrequent malformed polygon, possibly as little as 1 per hour. So I'm sure anyone reading this would think its obviously my GFX card, but other weirdness has put doubt in my mind.

firstly pixel bleeding (is that the appropriate term?), this is even present in the BIOS (although there are other issues there too which I'll get to late).


I'm also getting very weird (and they have to be faulty, as the heat sink for the mobo is just warm to the touch, certainly its not nearly 200 degrees c)

I've since installed the ASUS PC Probe and it shows the motherboard running at 38 and the CPU at 35.

Also as i said earlier, in the BIOS there are these pixel errors but also certain letters are corrupted there too (I've tried to flash the bios dozens of times, the confirmation check fails at 87% every time, no matter which version i try). "Quit" is spelled "Qui4" and "Disabled" is replaced with "Disable"". There is also some coloring issues in the BIOS which has some character backgrounds orange rather than blue, or the character itself colored wrongly.

I've also tried under clocking a few components to see if this would fix anything with no success, although i left the ram under clocked as the sticker on the ram says 800mhz 4-4-4-12 timing yet the SPD says 666mhz with 5-5-5-15 timings... I've gone with the SPD until i get the rest of this fixed, although it ran fine for a year at 4-4-4-12 etc.

My system is getting to a point that its almost unplayable

I think that's about everything so I'll list my components.
____________________________________

AMD Athlon 64 X2 4600+
ASUS M2N-SLI DELUXE
GIeL DDR2 800mhz 1gig x2
3D Fusion GeForce 7600 GT
Cooler Master eXtreme Power Duo 650w
Maxtor 6Y090M0 80gig
______________________________________

Most of it is a little over a year old, the HD is a little older as i bought it for a previous system but it only supported SATA as a raid.

I'm trying to hold out on getting a gfx card as long as possible with the 8900's so close I'm hoping to save some money there. I also do want something that supports DX10 with hardware as Age of Conan is something I'm really looking forward to.

Any and all advice appreciated. :) 
January 15, 2008 10:32:55 AM

I have roughly the same problem with a 7900 gt.Even though it has a decent cooler(EVGA KO),as soon as the card reaches a certain temp range artifacts like you showed start cropping up.Keeping case temps down and upping the fan speed on the card in 3d mode with rivatuner seems to delay or prevent the onset of said artifacting.So other than better cooling solutions(case,card or both)I am looking at newer cards.
January 15, 2008 11:15:25 AM

Well if it does it in BIOS to then maybe its your mobo, DVI cable, or maybe comunication between the mobo and the graphics card. This program that I'm linking you to can check all hardware for errors, be sure to uncheck any hardware that you don't have. Good luck!

Run this program - Burnin test 5.3 >>>> www.passmark.com
Related resources
January 15, 2008 11:40:30 AM

Something that is very easy to check is if your heat sink and fan on your graphics card (and cpu while you're at it) are full of dust. My friend was having that exact same problem a week ago and he could not figure it out. I came over and looked at his case (full of dust) and immediately thought to look at his heat sink. Sure enough it was FULL of dust to the point there was almost no air flow. We cleaned that out thorougly and applied some arctic silver5 and now hes off to the races. His card was a 7800GT from EVGA. He has been running for a week now with none of the symptoms as before and everything runs smooth once again as it used to.

Just a thought
January 15, 2008 11:48:45 AM

Look at the components on the video card, I just sent back a video card for caps that popped. It was a 7600 GT too.
Go to bad caps .com and check it out.
200 c is closing in on the flash point of paper.
http://downloads.guru3d.com/download.php?det=163
get rivatuner and let us know what your video temp is.
January 15, 2008 1:05:41 PM

I already stripped the system and give it a full dusting hoovering etc. Made sure to brush out the heat sink and fans to remove the persistent bits.

Rivatuner shows GPU core at a stable 45 degree.

I've also checked to make sure the heatsink on both the motherboard and gfx card are secure, and they are.

running BurnInTest now
January 15, 2008 3:21:29 PM



I then ran it again with just 3d graphics test and a full screen window... that test passed. before that i tried 4 windows, all 1280x1024... my PC crashed.

I'll run it again to try and find the right setting to push the gfx card, but not to breaking point.

are there any other tests that i can run? I'd like to know if the gfx card is the only thing at fault here before i move on.

I had hoped this system would last another few months for chipset revisions that i think are due later in the year...
January 15, 2008 8:49:14 PM

You just got 15,000 errors in the 2D test, its your videocard thats the problem. You said yourself that it happens in BIOS also which is 2D, your still using your videocard when in BIOS. Why would you want to run another test when you just failed a 2D test? :heink: 

Video memory corruption + 15,000 errors = bad videocard memory.
January 15, 2008 9:28:58 PM

Surely the BIOS doesn't use a video card to decide if it should print and t or a 4 on the screen.

That's mainly why I'd like to check.

my choice of video card will depend on if this mobo has a future or not.
January 15, 2008 9:39:27 PM

graphics cards die, my 7900gs was doing the same thing, so i replaced it.
January 16, 2008 5:38:55 AM

xanoth said:
Surely the BIOS doesn't use a video card to decide if it should print and t or a 4 on the screen.



I never said that it did, now did I? Without the videocard you wouldn't be able to see anything but a blank black screen.

BTW your welcome!
January 16, 2008 7:33:38 AM

The initial advice was much appreciated, the sarcasm later wasn't.
January 16, 2008 10:20:04 AM

Bad Ram (atleast ive seen on ONBOARD gpus's) will corrupt text and display strange characters, i got a few smiley faces in the bios once :) 
January 16, 2008 11:41:00 AM

All BIOS should have smiley faces in :) 

I've just ordered a 8800GTX (rationalized it as a late Christmas present to myself)

Thanks for all the advice, help and suggestions, I certainly have more confidence now that i won't have to buy a new mobo soon also. :D 
!