Contact IT Support

 

 

 

 

 

 

CUDA

CUDA

CUDA is NVIDIA’s parallel computing architecture that enables dramatic increases in computing performance by harnessing the power of the GPU (graphics processing unit). The broad-ranging uses for CUDA includes image and video processing, computational biology and chemistry, fluid dynamics simulation, CT image reconstruction, seismic analysis, ray tracing, and much more.
Support site:
https://developer.nvidia.com/cuda-toolkit-40
Eligibility
Alumni
Faculty
Post-Docs
Students
License Type
Freeware
Installation instructions

Linux installation instructions

Use modules to load the applications available in the scientific application stack.
Please contact IT Linux Support or open a ticket with the IT Help Desk to install a new version of this application.

Loading CUDA from the scientific application stack

To run Cuda, first load the associated module. For example to load the latest version of Cuda execute:

 

$ module load cuda

 

Then execute either of these commands:

CUDA Compiler driver:

$ nvcc --version

OR CUDA debugger:

$ cuda-gdb