Using nd_range with parallel_for does not work

Hi,

it’s not quite clear what might be the problem: compile_program_error indicates there might be something wrong with the kernel, but the kernel is empty. I’d recommend adding error checking, you can look at some ComputeCpp SDK samples to see how that’s done.

Additionally, you can try enabling verbose output to get more information from the ComputeCpp runtime, see for example this post on how to enable it.

Have you observed different behavior depending on the ComputeCpp version used? Which version are you using at the moment?