Tesla on Asus P5N-E SLI?

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi,

I’m trying to get the Tesla card running on a Asus P5N-E SLI (nForce 650i, Intel Core 2 Quad Q6600) in combination with a 8800GTS in the blue PCI-E slot to drive the display.

I installed the latest drivers and got the deviceQuery from the SDK running. Unfortunately only the GTS is detected as CUDA capable device. Neither in the Windows XP device manager, nor under linux (/sbin/lspci), the existence of the Tesla board is visible (the device has power and its fan is running).

Does anybody have experience with this board and the Tesla card?

The motherboards manual says that the PCI-E goes from 16x to 8x when using a second card. Might this be a problem for the Tesla?

Does anybody know of a Tesla configuration with a Intel 775? Does the Asus P5N32-E SLI (nForce 680i SLI) work with the Tesla card?

Thanks for advice!

Regards,
Michael

2

If the Tesla card isn’t showing up on the PCI bus in either OS then this is most likely a motherboard problem (SBIOS or elsewhere). Of course if possible you should test whether the card works in any other motherboard.

3

Thanks for the help…

I got the Tesla running on this board now (with RHEL5). It seems like no card (GTS, Tesla) is detected in the secondary slot. Putting the Tesla in the primary slot and a graphics card in a PCI slot gives me access to the Tesla.

Regards,
Michael

4

Turning the little PCB on the motherboard to enable “SLI/Secondary Graphics Card” allows the MOBO to detect the card in the secondary PCI-E, even if it’s a Tesla and SLI itself is not used.