CUDA installation errors armhf, sm_10

System: Intel i7 running Ubuntu 12.04
Graphics Card: specifications from deviceQuery

Device 0: "GeForce GTX 650"
  CUDA Driver Version / Runtime Version          6.0 / 6.0
  CUDA Capability Major/Minor version number:    3.0
  Total amount of global memory:                 2048 MBytes (2147287040 bytes)
  ( 2) Multiprocessors, (192) CUDA Cores/MP:     384 CUDA Cores
  GPU Clock rate:                                1058 MHz (1.06 GHz)
  Memory Clock rate:                             2500 Mhz

I recently started working with CUDA. I am having the following issues:

  1. When I compile the codes I get the following warning
nvcc warning : The 'compute_10' and 'sm_10' architectures are deprecated, and may be removed in a future release.

I don’t understand this warning. Looking up online I found that compiling with


flag would not show the warning. It works, however, I don’t understand what is happening. It would be helpful if someone could explain. Is there a permanent solution?
Also, while installing CUDA, I had run

sudo sh -c   'echo "foreign-architecture armhf" >> /etc/dpkg/dpkg.cfg.d/multiarch'

It was unsuccessful however, the CUDA still got installed and works fine. Is this causing issue? How can I revert it? Also, when I do

sudo apt-get update

in my system, I get a huge amount of errors pertaining to armhf.

The compiler defaults to an sm_10 build target. As the warning tells you, the sm_10 build target is deprecated. Deprecated features are still fully supported in the current version, but that support will be removed in a subsequent version. At that point these warning messages will disappear. Instead you would see an error about an unsupported architecture. The warnings are provided to make CUDA programmers aware about the intention to remove those build targets well before support ceases.

You would want to chose your build target based on the compute capability of your GPU, which is 3.0. So you would want to build with -arch=sm_30.

Yes. Works good.