Hello and good morning External Image
I have a problem with the cuda engine External Image I have three devices on my x58 motherboard and the nvidia controll panel see all of them,but if i will use n applikation as tmpgenc are only the first 2 are available.
Sli is deaktivated!!
On the pictures you can see what i mean :ph34r:
Have anybody n idea what the problem is ?
Thank you already for your fixes External Image
Sorry for the two posts are i have no idea how i can clean one of them :blink:
Have you read the release notes? In Vista, you either have to have a monitor attached to each GPU or enable PhysX on that GPU in order for it to be usable by CUDA.
But why i can use two of them devices? Only in one card is a monitor plugt in?
This is not the badaboom converter its tmpeggenc,and in other cuda application is it exactly the same.
Look her only 0 and 1
Only two from three devices are visible in the programms
If there is the problem with the monitor,comes anytime n update to fix this monitor plug in problem?
Because i will not need three displays, i will only use this cards with cuda External Image
Nobody has another approach? External Image
Is this the only way “use 3 displays to work with 3 devices under cuda” ?
dose windows see all three cards ? can you extend your desktop to the other devices without attaching a monitor ? i have done this in the past (on windows xp) to solve some issues, just create the smallest screen possible.
Yes all cards a visible in vista.
Yes and i can extendet the desktop to other devices when i install n display on this device,but this is not the question.
I will not to install 1 display on each device to run cuda application on it :unsure:
And why the applikacations see two devices?
well then i guess you won’t be able to use all 3 devices with cuda under vista :) defining small desktops on the devices should have minimal effects on performance and your work flow …
Yes all three in different applications for converting External Image
But i have not 3 displays :unsure:
And the other thing is why cuda see two devices?
Only one device have a monitor and the same with the desktop