![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
I have been experiencing system lockups while playing wmv, avi, mpg files after adding a second GTX and enabling SLI mode. If I disable SLI, I am unable to reproduce the lockup. It usually occurs soon after switching to and from full screen mode, or while seeking ahead or back in the movie file. I have been in extensive contact with Evga, but have so far been unable to pinpoint the root cause.
I have done exhaustive memory testing, including downclocking my memory from 800 to 667mhz, loosening timings, tweaking memory voltages. I tried replacing my 4x1gb of crucial to 2x2gb of corsair, and still experienced the problem. Under all circumstances I have had hours and hours and hours of clean memtest86 tests and hours and hours of prime95 with no errors, including 10+ hours of prime95 with geoforms running in the foreground to stress the video system.
I have run both cards independantly in both slots and cannot distinguish between them in terms of stability or performance, they both run absolutely fine independantly, including 10+ hours each of looped 3dmark06 with everything maxxed at 2560x1600. Together in SLI mode I have run 3dmark06 looped for 15 hours with no lockups or errors, and any lockups in 3d games have been very uncommon on the order of maybe once per month.
I have tried many different individual codecs, divx, xvid, ffdshow, and even some of the codec packages such as k-lite. None of them resolves the problem. The lockups can occur with wmv files which require no 3rd party codecs at all. I uninstalled powerdvd and nero since they installed codecs and filters, but this did not help. Have tried 163.69, 169.09, 169.25, 169.28, 169.32, 169.38 drivers so far with none of them fixing it.
My PSU should be adequate, with a meter I get 12.06-12.08 volts on the 12v rail with little or no fluctuation, and a maximum of 16mV ripple under full load, with 2mV ripple idle. Temps are 75-77 degrees under load on the GPU's, 40ish on the cpu, 50-55 system, and are usually lower while reproducing the problem since WMP doesn't exactly stress most of those components much.
I am not overclocking, everything is running at stock speeds and voltages. System is as follows:
Vista ultimate x64
2xEVGA 8800GTX
Striker Extreme (tried BIOS 1303, 1305, 1401)
Core 2 Quad qx6700
4x1GB Crucial Ballistix 4-4-4-12 2.2v memory (was running it at 2.2 for a long time, now at 2.0, no change in stability or memtest performance)
Areca 1210 raid card
4x74GB Raptors (older model with the smaller caches)
1x160gb Seagate barracuda on internal sata
LG dvd-rw
PC Pwr&Cooling silencer 750
Dell 3007wfp-hc
Any suggestions would be appreciated, a handful of people have reported similar problems on the nvidia and slizone forums but not seen anyone with a resolution. Been at it for weeks now with evga support trying to figure out a solution, and basically gone nowhere.
I have done exhaustive memory testing, including downclocking my memory from 800 to 667mhz, loosening timings, tweaking memory voltages. I tried replacing my 4x1gb of crucial to 2x2gb of corsair, and still experienced the problem. Under all circumstances I have had hours and hours and hours of clean memtest86 tests and hours and hours of prime95 with no errors, including 10+ hours of prime95 with geoforms running in the foreground to stress the video system.
I have run both cards independantly in both slots and cannot distinguish between them in terms of stability or performance, they both run absolutely fine independantly, including 10+ hours each of looped 3dmark06 with everything maxxed at 2560x1600. Together in SLI mode I have run 3dmark06 looped for 15 hours with no lockups or errors, and any lockups in 3d games have been very uncommon on the order of maybe once per month.
I have tried many different individual codecs, divx, xvid, ffdshow, and even some of the codec packages such as k-lite. None of them resolves the problem. The lockups can occur with wmv files which require no 3rd party codecs at all. I uninstalled powerdvd and nero since they installed codecs and filters, but this did not help. Have tried 163.69, 169.09, 169.25, 169.28, 169.32, 169.38 drivers so far with none of them fixing it.
My PSU should be adequate, with a meter I get 12.06-12.08 volts on the 12v rail with little or no fluctuation, and a maximum of 16mV ripple under full load, with 2mV ripple idle. Temps are 75-77 degrees under load on the GPU's, 40ish on the cpu, 50-55 system, and are usually lower while reproducing the problem since WMP doesn't exactly stress most of those components much.
I am not overclocking, everything is running at stock speeds and voltages. System is as follows:
Vista ultimate x64
2xEVGA 8800GTX
Striker Extreme (tried BIOS 1303, 1305, 1401)
Core 2 Quad qx6700
4x1GB Crucial Ballistix 4-4-4-12 2.2v memory (was running it at 2.2 for a long time, now at 2.0, no change in stability or memtest performance)
Areca 1210 raid card
4x74GB Raptors (older model with the smaller caches)
1x160gb Seagate barracuda on internal sata
LG dvd-rw
PC Pwr&Cooling silencer 750
Dell 3007wfp-hc
Any suggestions would be appreciated, a handful of people have reported similar problems on the nvidia and slizone forums but not seen anyone with a resolution. Been at it for weeks now with evga support trying to figure out a solution, and basically gone nowhere.