C/C++ and Fortran

C/C++ and Fortran

Connect, learn, share, and engage with IBM Power.

 View Only

XL Fortran for Linux, V15.1.4 provides CUDA Fortran support

By Archive User posted Tue July 05, 2016 04:23 AM

  

Originally posted by: ChristineLi


IBM® XL Fortran for Linux, V15.1.4 supports the CUDA Fortran programming model, which is a subset of CUDA constructs, to exploit the NVIDIA GPUs. You can use the commonly used subset of CUDA Fortran that is provided by IBM XL Fortran for Linux to offload computations to the NVIDIA GPUs on the selected hardware and operating systems.

You can use the newly introduced xlcuf invocation command to enable CUDA Fortran support. Files with new file name extension, .cuf or .CUF, are also recognized as Fortran source files. The following compiler options are useful for CUDA Fortran programs.
-qcuda
The -qcuda option is added to enable the compiler support for CUDA Fortran.
-qcudaerr
The -qcudaerr option is added to control whether the compiler automatically inserts error checking code for CUDA API calls.
-W, -X
The -W option is updated to pass one or more options to specific compiler components, including the components that are used during compilation of CUDA Fortran programs. The -X option is added to pass one option to a specific compiler component , which can be any component that is used during compilation of CUDA Fortran programs.

For more information, see the Getting Started with CUDA Fortran programming using XL Fortran.

0 comments
3 views

Permalink

https://community.ibm.com/community/user/blogs/archive-user/2016/07/05/xl-fortran-for-linux-v1514-provides-cuda-fortran-support