-
Why are there different implementations for different backends in this example? |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 2 replies
-
They were written when we were experimenting with various GPU strategies. |
Beta Was this translation helpful? Give feedback.
-
Hi @lwJi, This tutorial predates the |
Beta Was this translation helpful? Give feedback.
Hi @lwJi,
This tutorial predates the
amrex::ParallelFor
functionality. Originally we were exploring different options for GPU support. However, we kept the OpenMP and OpenACC examples in place so people could have examples of using those, since they are still compatible with AMReX. The "CUDA" example shows our preferred way of doing things.