CUDA FORTRAN __launch_bounds__()

Is there an equivalent of launch_bounds() in CUDA FORTRAN? Please provide a usage example if there is.

Hi frnkyl004,

Not yet, though we do have an open feature request for it (TPR#19302). I’ll add you to the list of requesters.

  • Mat

Hey Mat

Could you add me to that list as well? Thanks.




Some simple support for launch bounds is available in 2020 compilers. Here is the syntax:

attributes(global) launch_bounds(256,8) subroutine test(a,b,c,n)