I currently own an Inno3D Geforce 580 GTX (release version) and am experiencing issues with DX10 and DX11 games.
Games running under DX10 or 11 randomly freeze with green artifacts covering the screen. The display driver displays an error message stating that it stopped working and reset, and the system is operable thereafter, however this has persisted since November 2010 (running 263.xx) all the way until today (running 275.xx).
Note: The card is stock clocked, idles at ~38 degrees and maxes out ar ~80 degrees during metro2033/crysis 2 dx11/witcher 2 and around 68 - 72 degrees on less taxing games (left 4 dead2, starcraft 2, dead space 2, fear 2 etc).
When this occured while playing Crysis in DX11 last November, forum goers speculated that it was most likely a driver issue, but 7 months down the line this simply cannot be the case.
This occurs in Crysis (dx10), FEAR 3 (dx11) and Crysis 2 (dx11).
I have never experienced this issue in DX9 games, which includes the games above, and up until now all of the games i have been playing have been DX9 based (or with dx10/dx11 disabled). I did play Metro 2033 for a while in both DX10 and DX11 and did not experience artifacts, and if The Witcher 2 incorporates DX11 i have not experienced artifacts in that game either yet.
I have tried another rail and have reseated several times, the issue persisted.
This is also not an isolated matter, a lot of 580 owners are experiencing the same issue, but no real cause has been identified. Overcvlocked cards were downclocked to stock as a quickfix which did solve the issue, but mine i s a stock card and i refuse to downclock.
Thanks for the help, i've RMA'd the card. Fingers crossed.
August 10, 2011 8:39:42 AM
and ? did the RMA solved a problem ? btw I have same issue.. exactly same.. i have saw on some forums that it might be overheating VRam. and it givs logic to me since stock GTX 580 do not have the best cooler for a heat it makes and you mentioned that some overclockers solved it by lowering their clocks to stock. and it all depends on the airflow you have in your case as well. in my case i will try to make some home made cooling method and i will let you know then.
It is either vram or the power vrm as the temp for the core is well with in the limits and much of the image is clean so it has to be one of the vram modules being defective or the power vrm not providing enough stable current.
I have had the exact same problem since I bought my GTX580. I replaced my Corsair HX520W with a Corsair AX850W because someone suggested that the PSU I had was not good enough.
But that did not work so I RMA'd the card, they ran test and things like that but found nothing wrong at all.
So I got the same card back and it is crashing in some games... Please help me I don't know what to do.