Compatability query for Dell 1950 and Tesla S

Hi All,

We’re interested in getting a Tesla S series to run in conjunction with a linux host, and rather than splash out on a new box we’re keen to use an existing machine if possible. The candidate we have in mind is a Dell 1950 PowerEdge server, and dmidecode indicates that we have two PCI Express x8 slots available:

Handle 0x0900, DMI type 9, 13 bytes
System Slot Information
Designation: PCI1
Type: x8 PCI Express
Current Usage: Available
Length: Long
Characteristics:
3.3 V is provided
PME signal is supported

Handle 0x0901, DMI type 9, 13 bytes
System Slot Information
Designation: PCI2
Type: x8 PCI Express
Current Usage: Available
Length: Long
Characteristics:
3.3 V is provided
PME signal is supported

So, can anyone vouch for this platform? It will run Debian “lenny”, which I have working fine with CUDA on a GeForce 9600 GT.

Thanks for any help.

Hmm, maybe I should have spent some more time delving through NVIDIA’s CUDA webpages. So, I’ll answer my own question. I notice that on [url=“http://www.nvidia.com/object/tesla_compatible_platforms.html”]http://www.nvidia.com/object/tesla_compatible_platforms.html[/url] it states for the S870:

Vendor Model Config Options(where appropriate) Status

Dell PowerEdge 1950 III Can Support one or two PCIe connections to Tesla 1U. Requires x8 PCIe host adapter card Check mark**

** Passes all functional testing. Mechanical interference with chassis cover.

This seems quite positive, but does anyone know just what level of “mechanical interference” we’re talking about here? Also, I can’t find a compatability table for the Tesla 10 series. Will it be equally compatible?

Thanks for any help.