The High Performance Computing (HPC) folk are taking to Python in droves. They are not deserting Fortran and C++, they are still the languages in which the computation codes are written. Python is though the coordination language. Also Python is the language of writing user interfaces and performing data visualization.
ctypes, SWIG, Boost.Python, and Cython all have roles to play in connecting Python to C++ and Fortran codes, making Python a far more flexible tool than Maple, Mathematica, etc. for result processing and data visualization. And SciPy rocks.
But this is only one aspect of high performance. Traditionally, people view bytecode/virtual machine-based languages as slow and clunky. Java put paid to that idea with its just-in-time compilation. Python is similarly a bytecode/virtual machine-based language but until recently had not been seen as being in need of being a high performance language. This is changing.
This free talk from Dr Russel Winder will take an interactive wander through all these issues, resulting in the conclusion that Python is up there with C++ and Fortran as a high performance language.
Official Website: http://skillsmatter.com/podcast/home/high-performance-python/js-1641
Added by skills.matter on November 9, 2011