Thats not a very good photo we can't see anything.
This is a family kind of place!
0
Mt_GoatHead Cheezy KnobPflugerville (north of Austin)Icrontian
edited January 2009
I finally got back to it again. So far no go. So far I have;
checked all the things listed in this thread so far!
uninstalled, rebooted and reinstalled CUDA DRIVERS
hooked up a display to the second card to see if that changed anything (stranger thigs have worked b4) NEVER COULD GET DISPLAY TO WORK ON SECOND CARD/ either as second display or solo!
switched cards in slots/ no difference, did verify that both cards wok just fine
switched target in shorcut boxes/ changed GPU1 box target from -GPU 0 to -GPU 1, got same error as GPU2 box was getting all along, changed GPU2 target box from -GPU 1 to GPU 0 and it worked just fine
switched jumpers for 16x PCI express slots from 16x and 2x to 8x and 8x/ did not make second card work, first card foled slower
tried SLI to see if anything changed/ did verify that SLI worked, but did not meet my expectations
I am currently removing all chipset and video drivers then going to reinstall and try again.
0
Mt_GoatHead Cheezy KnobPflugerville (north of Austin)Icrontian
edited January 2009
Did the reinstall of system and video drivers and still the same. ;( Also tried different drivers with no luck. At this point I am ready to give it up! This sure has been playing hell with my points production!!!
0
LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited January 2009
One other thing you could try:
For the second folding client, in the start up, set the flag "-forcegpu nvidia_g80" (space between ~u and g~). This was used quite a bit only multi clients with older drivers, but you could try now as well. Note: yes, G80, even though your GPUs are all G92 core. It's just a signal to tell Windows that there is another Nvidia GPU that must be tasked separately from the primary GPU.
So assuming the startup for the FIRST GPU is designate "-gpu 0," then the second GPU-client-startup will be "-gpu 1 -forcegpu nvidia_g80."
0
Mt_GoatHead Cheezy KnobPflugerville (north of Austin)Icrontian
Comments
This is a family kind of place!
- checked all the things listed in this thread so far!
- uninstalled, rebooted and reinstalled CUDA DRIVERS
- hooked up a display to the second card to see if that changed anything (stranger thigs have worked b4) NEVER COULD GET DISPLAY TO WORK ON SECOND CARD/ either as second display or solo!
- switched cards in slots/ no difference, did verify that both cards wok just fine
- switched target in shorcut boxes/ changed GPU1 box target from -GPU 0 to -GPU 1, got same error as GPU2 box was getting all along, changed GPU2 target box from -GPU 1 to GPU 0 and it worked just fine
- switched jumpers for 16x PCI express slots from 16x and 2x to 8x and 8x/ did not make second card work, first card foled slower
- tried SLI to see if anything changed/ did verify that SLI worked, but did not meet my expectations
I am currently removing all chipset and video drivers then going to reinstall and try again.For the second folding client, in the start up, set the flag "-forcegpu nvidia_g80" (space between ~u and g~). This was used quite a bit only multi clients with older drivers, but you could try now as well. Note: yes, G80, even though your GPUs are all G92 core. It's just a signal to tell Windows that there is another Nvidia GPU that must be tasked separately from the primary GPU.
So assuming the startup for the FIRST GPU is designate "-gpu 0," then the second GPU-client-startup will be "-gpu 1 -forcegpu nvidia_g80."