[FEA]: Implement CUDA backend for parallel `cuda::std::find`

[miscco]

We want to use cuda::device_find_if to implement

• cuda::std::find
• cuda::std::find_if
• cuda::std::find_if_not