CUDA Toolkit and Compilers

Revision as of 19:40, 5 May 2017 by Tgreen (talk | contribs)

Description

From the NVIDIA CUDA Toolkit Home Page: The NVIDIA® CUDA® Toolkit provides a comprehensive development environment for C and C++ developers building GPU-accelerated applications. The CUDA Toolkit includes a compiler for NVIDIA GPUs, math libraries, and tools for debugging and optimizing the performance of your applications.

Version

  • 8.0

Authorized Users

  • CIRCE account holders
  • RRA account holders
  • SC account holders

Platforms

  • CIRCE cluster
  • RRA cluster
  • SC cluster

Modules

CUDA Toolkit and Compilers requires the following module file to run:

  • apps/cuda/8.0

Compiling with CUDA Toolkit on CIRCE/SC

The CUDA Toolkit and Compilers user guide is essential to understanding the application and making the most of it. The guide and this page should help you to get started with your simulations. Please refer to the Documentation section for a link to the guide.

  • Note on CIRCE: Make sure to run your jobs from your $WORK directory!

Running CUDA Toolkit and Compilers on CIRCE/SC

The CUDA Toolkit and Compilers user guide is essential to understanding the application and making the most of it. The guide and this page should help you to get started with your simulations. Please refer to the Documentation section for a link to the guide.

  • Note on CIRCE: Make sure to run your jobs from your $WORK directory!
  • Note: Scripts are provided as examples only. Your SLURM executables, tools, and options may vary from the example below. For help on submitting jobs to the queue, see our SLURM User’s Guide.
  • Note: All CUDA jobs should be run in an SBATCH/SRUN session on the cuda partition, and NOT on the login nodes!

To run CUDA tools, the following steps must be taken, in addition to using another command sequence.

1. To start an Interactive SRUN session, run the command below with example resources:

[user@login0 ~]$ srun --time=02:00:00 --nodes=1 --ntasks=4 --gres=gpu:1 --pty /bin/bash
[user@wh-520-9-2 ~]$ 

 

2. Next, load the CUDA module as described above, and then start your specific CUDA tool(s):

[user@wh-520-9-2 ~]$  module load apps/cuda/8.0

 

3. Finally, run your CUDA-based code or binaries.

Documentation

Home Page, User Guides, and Manuals

More Job Information

See the following for more detailed job submission information:

Reporting Bugs

Report bugs with CUDA Toolkit and Compilers to the IT Help Desk: rc-help@usf.edu