As we general gain on flexibility, debugability and explainability compared to fortran/c++ solutions, it would be good to know how much speed offer for this.
For this, we should first look at good fortran/cpp solution of e.g. a Random Walk and what exactly they do implement, fix different size and other constrains for the simulation (different hardware, too) before we measure both solutions.
Furthermore, configuring a GPU-executer for speed-comparison would be nice to have.
[ ] find a nice (simple, still not rivial, good quality) fortran/cpp simulation
[ ] define different hardware and test sizes for comparism
As we general gain on flexibility, debugability and explainability compared to fortran/c++ solutions, it would be good to know how much speed offer for this.
For this, we should first look at good fortran/cpp solution of e.g. a Random Walk and what exactly they do implement, fix different size and other constrains for the simulation (different hardware, too) before we measure both solutions.
Furthermore, configuring a GPU-executer for speed-comparison would be nice to have.