I want to optimize the performance of the boolean operation of boost::polygon with CUDA.

I  want to speedup it at the source code level, and I found it is hard to speedup the hot spot because of dependency & container.

Then, I  want to evaluate the feasibility from the algorithm. I have found the article "Industrial strength polygon clipping - A novel algorithm with applications in VLSI CAD", but I think it is hard to understand.

Do you have more material for study?


Thanks,

Rogy