CUDA Tesla M60 GPU Not Detected

Hello, first post, not sure if this is the correct place to ask this, been using Quadros with a CUDA app for a long time and we are testing our app with ESXi and Tesla M60 but it doesn’t seem to be detected:

C:\cudatest\cdpsimpleprint.exe
starting Simple Print (CUDA Dynamic Parallelism)
CUDA error at C:/ProgramData/NVIDIA Corporation/CUDA Samples/v8.0/0_Simple/cdpSi
mplePrint/cdpSimplePrint.cu:134 code=38(cudaErrorNoDevice) “cudaGetDeviceCount(&
device_count)”

C:\cudatest>test_cudacheck.exe
cudaGetDeviceCount returned 38
→ no CUDA-capable device is detected
Result = FAIL
cuda check status: 0

I am trying to use two .exe to verify CUDA is installed because our app claims its not – the IT who manage the server also says it is fine and the VM has the Driver installed (Windows Server 2012 R2). They said they use the same driver on other VMs as its from VMWare for their ESXI version. Any thoughts? dxdiag detects both VMWware 3D virtual GPU and Telsa M60.

Is this a GRID setup? Are you using GRID to create a GPU in the VM? Or are you using PCI passthrough?

If you are using GRID, currently, its necessary to specify a maximal GRID profile in order to have CUDA available. Other GRID profiles do not support CUDA.

If you are using VMWare VSGA graphics, that doesn’t support CUDA either.

if you are using GRID it’s probably better that you post on the GRID forums.