NVIDIA GPU Computing SDK
The NVIDIA GPU Computing SDK provides various examples in CUDA C (and in older versions also examples in OpenCL C). They can be used to verify the correct setup of the GPU, as starting point for your own application and to give you an idea how to implement certain algorithms on a GPU.
The CUDA samples come with the CUDA toolkits. Thus, they can be found in the cluster's toolkit installation path (denoted in the environment variable $CUDA_ROOT). Since you will have to build the examples, you need write permissions. Therefore, copy the samples directory to your home or work directory (by following symbolic links):
cp -r -L $CUDA_ROOT/samples $WORK/.
Then go to samples and make all CUDA examples by denoting the installation path.
If you would like to build only selected examples, you can find the source codes in the samples folder as well. In either case you will find the executables in bin -> linux -> release.