No device supporting CUDA was found

Running matlab and entering:

>> gx = gpuArray(rand(2^16,1))
Error using gpuArray (line 28)
No device supporting CUDA was found.

I have a unix system with

Red Hat Enterprise Linux Server release 6.3 (Santiago)

Matlab R2011b

/etc/X11/xorg.conf
Section "Device"
        Identifier  "Device0"
        Driver      "nvidia"
        VendorName  "NVIDIA Corporation"
EndSection

Cuda is installed in /usr/local
Cuda Language Front End
Portions Copyright (c) 2005-2012 NVIDIA Corporation
Portions Copyright (c) 1988-2012 Edison Design Group Inc.
Based on Edison Design Group C/C++ Front End, version 4.3.1 (Apr  5 2012 00:25:22)
Cuda compilation tools, release 4.2, V0.2.1221

How do I find the device to support CUDA?

   nvidia-smi --query
  FATAL: Error inserting nvidia (/lib/modules/2.6.32-279.5.2.el6.x86_64/weak-updates/nvidia/nvidia.ko): No such device
  NVIDIA: failed to load the NVIDIA kernel module.
  Nvidia-smi has failed because it couldn't communicate with NVIDIA driver. Make sure that latest NVIDIA driver is installed and running.

Do I need to install a certain NVIDIA driver?

Answer 1

[Jason Ross]

You need to download the most recent CUDA driver from nVidia's site and install it. As of right now, that's 304.51.

http://www.nvidia.com/Download/index.aspx?lang=en-us

Also, what GPU are you running? It needs to be Compute Capability 1.3 or higher:

http://developer.nvidia.com/cuda/cuda-gpus

From the error message above, it looks like the CUDA toolkit/SDK is installed, but to use it, you also need to install the CUDA driver itself on the system. The nvidia.ko file is the driver that that system needs to have loaded to access the GPU.

When things are installed correctly, nvidia-smi should list all the GPUs in the system and a lot of information about their current state (fan speed, temp, etc)