Information about software packages installed on the ICHEC systems.
nVidia CUDA compiler and SDK
fionn: 5.0 / 5.5 / 6.5
CUDA is NVIDIA’s parallel computing architecture that enables dramatic increases in computing performance by harnessing the power of the GPU (graphics processing unit).
Those CUDA compilers and SDK are meant to allow GPGPU computing in the GPU enabled partition of the machine. If you want to get access to this facility, please open a helpdesk call.
Sample codes are provided with the SDK and compiled for your convenience. To get access to the corresponding binaries, just use the "-samples" module:
The sample folder containing all the corresponding sources is located into the CUDA installation directory.