Warp divergence in independent thread scheduling?

Accelerated Computing CUDA CUDA Programming and Performance
2
  1. Yes. All GPUs of Volta family or newer have the volta thread execution model (independent thread scheduling).
  2. It is always active, you cannot disable it. (You might be able to disable it for Volta architecture on some CUDA versions, by compiling for an arch less than 7.0, but this is something I would not rely on, and it would limit you from doing the right thing in terms of compilation strategy).
  3. Warp divergence may still have a cost.
  4. Here is an example.