Rig:
CPU: 3500+ Athalon 64
Mobo: A8N-SLI Deluxe
RAM: 1 gig Samsung DDR400
Video: Gigabyte 6600GT Turboforce 128 PCI-E (NX66T128VP)
HDD: Maxtor 160 SATA
Case: Thermaltake Xaser III Viking with Thermaltake 420W PSU
Drivers:
Direct-X 9.0c
nVidia 71.89 display drivers
nForce4 Standalone Kit 6.53
Description:
Since the last patch in WoW, I'm having some very serious texture tearing problems. Never had these problems before the patch, but suddenly torn textures appear in explosion type patterns all over the screen (sometimes these torn textures have strings of text within them). So, I post about it in WoW tech forums.
Someone mentions it's heat related -- so I start monitoring my heat levels (under the nVidia driver tab in display properties). Now that I am looking, I find that my usual temp is never less than 100 degrees C. Typically its 110+ just standing around in the game, without really doing anything. When I actually do play the game, it spikes at 125+ (highest I have seen it go is around 133). Needless to say, it's when I have these high temps that I get the tearing. A friend bought the exact same rig (that I built) and his normal temp is about 80 degrees C after hours of playing. I find that these temps can peak very quickly from cool to really hot.
Some days my temp doesn't peak however, and I can play normally all night without too many problems. When the tearing appears I keep playing and the tearing clears itself and comes back infrequently. Some days it peaks really high, like today -- was getting over 120+ C just by having WoW running in Darnassus. Our rigs are less than a month old.
I've tried underclocking my RAM, changing the 1T/2T, changing the Hypertransport rate in my BIOS, as well as toggling all in-game shaders for texturing, and changing the texturing multisampling rate. If I fool with the various settings (some of which reduce graphical performance, some of which boost it) I can produce between 60+ and ~80 FPS. I can't relate these settings back to the relative duration of my play, because I don't ah heck around with them on a daily basis -- so this can't be to blame (i.e., less playing time due to higher FPS').
Putting my case fans up to full blast sometimes helps, sometimes doesn't. The card in question doesn't have a core fan, but it does have two giant heatsinks with a bridge connecting them, and they are uncomfortable to the touch when checked while reading high temp (so I know they aren't unseated). The card comes overclocked out of the box to come close to matching the 6800GT benchmarks, which is why I bought the damn thing (then underclock it dummy! Right well, I haven't overclocked it since I got it -- this is occurring with bare minimum settings).
My board sensor never passes 40 degrees. I have checked the CPU temps in Sandra and they are low as a general rule as well. I've tried popping off the side of my case, but this doesn't seem to help much (about as much as turning the case fans up).
I should also mention that changing nForce or video drivers to earlier versions tends to produce hard or soft lockups instead of the tearing, requiring either a hard boot or causing an immediate crash to desktop with a nebulous error message. Tried running in Open GL mode once too, but that eventually produced a massive tear that hard locked the machine. All DX Diag tests complete normally.
In short, I'm at a loss. I have no idea what my previous temps were, since I never checked them prior to experiencing this texturing problem (that said I know these temps are ridiculously high, I'm just wondering why all of a sudden).
Any ideas? Think I got a bum card? (if so why is it acting up post-patch? It's almost as if a switch has been thrown from "great" to "crap". Perhaps the patch changed something in the way textures are drawn that is causing this error? I suggest this because there are many other people over at WoW tech support with similiar problems. Some experience total texture drawing failure of e.g., water, sky or ground, or have transparency of background doodads. Some see all textures totally garbled or fuzzy, while other see patches of squares that don't match the background. However, WoW isn't singularly to blame -- the tearing also occurs in HL2. I'm also aware a lot of people over at nVidia hardware forums are experiencing the texture fuzziness with 6600-6800GTs).
My error looks exactly like the screenie of the OP in this thread:
http://forums.nvidia.com/index.php?showtopic=4696
...but more severe, with four times as many "rays." I find it difficult to capture the "rays" of garbled texture in screenies, as taking the screenie totally clears the effect for a short while.
So -- does the card get skipped along the pavement in the general direction of Gigabyte's mailbox?
Thanks for any ideas,
Ashur
CPU: 3500+ Athalon 64
Mobo: A8N-SLI Deluxe
RAM: 1 gig Samsung DDR400
Video: Gigabyte 6600GT Turboforce 128 PCI-E (NX66T128VP)
HDD: Maxtor 160 SATA
Case: Thermaltake Xaser III Viking with Thermaltake 420W PSU
Drivers:
Direct-X 9.0c
nVidia 71.89 display drivers
nForce4 Standalone Kit 6.53
Description:
Since the last patch in WoW, I'm having some very serious texture tearing problems. Never had these problems before the patch, but suddenly torn textures appear in explosion type patterns all over the screen (sometimes these torn textures have strings of text within them). So, I post about it in WoW tech forums.
Someone mentions it's heat related -- so I start monitoring my heat levels (under the nVidia driver tab in display properties). Now that I am looking, I find that my usual temp is never less than 100 degrees C. Typically its 110+ just standing around in the game, without really doing anything. When I actually do play the game, it spikes at 125+ (highest I have seen it go is around 133). Needless to say, it's when I have these high temps that I get the tearing. A friend bought the exact same rig (that I built) and his normal temp is about 80 degrees C after hours of playing. I find that these temps can peak very quickly from cool to really hot.
Some days my temp doesn't peak however, and I can play normally all night without too many problems. When the tearing appears I keep playing and the tearing clears itself and comes back infrequently. Some days it peaks really high, like today -- was getting over 120+ C just by having WoW running in Darnassus. Our rigs are less than a month old.
I've tried underclocking my RAM, changing the 1T/2T, changing the Hypertransport rate in my BIOS, as well as toggling all in-game shaders for texturing, and changing the texturing multisampling rate. If I fool with the various settings (some of which reduce graphical performance, some of which boost it) I can produce between 60+ and ~80 FPS. I can't relate these settings back to the relative duration of my play, because I don't ah heck around with them on a daily basis -- so this can't be to blame (i.e., less playing time due to higher FPS').
Putting my case fans up to full blast sometimes helps, sometimes doesn't. The card in question doesn't have a core fan, but it does have two giant heatsinks with a bridge connecting them, and they are uncomfortable to the touch when checked while reading high temp (so I know they aren't unseated). The card comes overclocked out of the box to come close to matching the 6800GT benchmarks, which is why I bought the damn thing (then underclock it dummy! Right well, I haven't overclocked it since I got it -- this is occurring with bare minimum settings).
My board sensor never passes 40 degrees. I have checked the CPU temps in Sandra and they are low as a general rule as well. I've tried popping off the side of my case, but this doesn't seem to help much (about as much as turning the case fans up).
I should also mention that changing nForce or video drivers to earlier versions tends to produce hard or soft lockups instead of the tearing, requiring either a hard boot or causing an immediate crash to desktop with a nebulous error message. Tried running in Open GL mode once too, but that eventually produced a massive tear that hard locked the machine. All DX Diag tests complete normally.
In short, I'm at a loss. I have no idea what my previous temps were, since I never checked them prior to experiencing this texturing problem (that said I know these temps are ridiculously high, I'm just wondering why all of a sudden).
Any ideas? Think I got a bum card? (if so why is it acting up post-patch? It's almost as if a switch has been thrown from "great" to "crap". Perhaps the patch changed something in the way textures are drawn that is causing this error? I suggest this because there are many other people over at WoW tech support with similiar problems. Some experience total texture drawing failure of e.g., water, sky or ground, or have transparency of background doodads. Some see all textures totally garbled or fuzzy, while other see patches of squares that don't match the background. However, WoW isn't singularly to blame -- the tearing also occurs in HL2. I'm also aware a lot of people over at nVidia hardware forums are experiencing the texture fuzziness with 6600-6800GTs).
My error looks exactly like the screenie of the OP in this thread:
http://forums.nvidia.com/index.php?showtopic=4696
...but more severe, with four times as many "rays." I find it difficult to capture the "rays" of garbled texture in screenies, as taking the screenie totally clears the effect for a short while.
So -- does the card get skipped along the pavement in the general direction of Gigabyte's mailbox?
Thanks for any ideas,
Ashur
Facebook
Twitter