I have multiple GPU’s and I don’t understand how I can tell the compiler which GPU to use when using OPENACC.
Compiler is not responsible for selecting GPU.
By default the first GPU is selected.
You can select required GPU by hands:
How do I know which one is which. I have a Tesla 2075 and Quadro 5000, and would like to use the Tesla for computations. Is there a way to set which one is the first one and see details…
Run the command “pgaccelinfo” to see which is which and the detail of each.