$customHeader
Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 13 Next »

Overview

This page is focused on NVidia’s HPC SDK. This SDK is currently under exploration by the ARCC team so please appreciate that support is limited but feel free to use and send any feedback to arcc-help@uwyo.edu.

The NVIDIA HPC Software Development Kit (SDK) is a comprehensive suite of compilers, libraries and tools for HPC, including the proven compilers, libraries and software tools essential to maximizing developer productivity and the performance and portability of HPC applications.

Using

Use the module name hpc-sdk to discover versions available and to load the application.

Documentation

22.7

20.7

Compilers:

  • nvc: C11 compiler for NVIDIA GPUs and Intel CPUs. It invokes the C compiler, assembler, and linker for the target processors with options derived from its command line arguments. Supports:

    • ISO C11

    • GPU programming with OpenACC

    • multicore CPU programming with OpenACC and OpenMP.

  • nvc++: C++17 compiler for NVIDIA GPUs and Intel CPUs. It invokes the C++ compiler, assembler, and linker for the target processors with options derived from its command line arguments. Supports:

    • ISO C++17

    • GPU programming with C++17 parallel algorithms (pSTL) and OpenACC

    • multicore CPU programming with OpenACC and OpenMP.

  • nvfortran: Fortran compiler for NVIDIA GPUs and Intel CPUs. It invokes the Fortran compiler, assembler, and linker for the target processors with options derived from its command line arguments. Supports:

    • ISO Fortran 2003 and many features of ISO Fortran 2008

    • GPU programming with CUDA Fortran and OpenACC

    • multicore CPU programming with OpenACC and OpenMP.

  • nvcc: CUDA C and CUDA C++ compiler driver for NVIDIA GPUs. nvcc accepts a range of conventional compiler options, such as for defining macros and include/library paths, and for steering the compilation process. Produces:

    • optimized code for NVIDIA GPUs and drives a supported host compiler for Intel CPUs.

Version 22.7, CUDA and GPUs:

Has been installed and uses CUDA 11.7 by default.

If you specifically need to use CUDA then load this module using:

module load cuda/11.7.1

and when creating an interactive session (salloc) or submitting a batch script (sbatch) then remember to request a GPU resource and use srun when running any executables. Further information of requesting GPUs on teton can be found here.

Version 20.7, CUDA and GPUs:

Has been installed and uses CUDA 10.1 by default.

If you specifically need to use CUDA then load this module using:

module load cuda/10.1.243

and when creating an interactive session (salloc) or submitting a batch script (sbatch) then remember to request a GPU resource and use srun when running any executables. Further information for requesting GPUs on teton can be found here.

Math Libraries

For further documentation see the respective pages:

22.7

20.7

Tools:

Both the Nsight_Compute and Nsight_Systems tools have been installed, but not tested. Documentation on their use will be added as we use/learn. For further documentation see:

22.7

20.7

  • No labels