Skip to content

Latest commit

 

History

History
10 lines (8 loc) · 866 Bytes

README.md

File metadata and controls

10 lines (8 loc) · 866 Bytes
Raw

cuda-examples

Examples of GPU acceleration with CUDA and OpenACC in C++ and Fortran.

Requirements

  • NVIDIA HPC SDK 23.1: These examples depend on different modules and libraries like CUDA, OpenACC, cuSPARSE and Thrust. These are included in the NVIDIA HPC SDKs and are easy to include/link with NVIDIA compilers.

Description of examples

  • SpMV_Fortran_OACC: UNDER DEVELOPMENT... will be an implementation of sparse matrix-vector multiplication using OpenACC directives in Fortran. TODO: Implement CUDA kernel to compute SpMV.
  • Thrust_interop: Shows how to couple Fortran using with OpenACC directives with Thrust library, by using interfaces with CUDA Fortran.
  • cuSPARES_Fortran_example: Shows how to perform sprase matrix multiplication on Fortran arrays ported to GPU using OpenACC, by creating a CUDA Fortran wrapper to cuSPARSE library.