[FEA]: Implement cuda.parallel version of cub::DoubleBuffer

[gevtushenko]

Is this a duplicate?

• I confirmed there appear to be no duplicate issues for this request and that I agree to the Code of Conduct

Area

Not sure

Is your feature request related to a problem? Please describe.

Sorting algorithms allocate a lot of memory. As a way to workaround this, we have cub::DoubleBuffer version of sorting API. Before we start working on cuda.parallel sorting algorithms, we need equivalent abstraction in Python.

Describe the solution you'd like

Implement cuda.parallel version of cub::DoubleBuffer.

Describe alternatives you've considered

No response

Additional context

No response