Skip to content

Consolidate vtkm::cont::cuda::internal classes

We currently have some VTK-m classes that make understanding the CUDA backend harder than needed.

Tasks:

  • Remove ArrayManagerExecutionThrustDevice.h and move logic into ArrayManagerExecutionCuda.h
  • Move DeviceAdapterAtomicArrayImplementation into a separate header
  • Remove DeviceAdapterAlgorithmThrust.h and move logic into DeviceAdapterAlgorithmCuda.h
Edited by Allison Vacanti