Functional programming and High Performance ComputingЯзык Python, его эволюция и использование
Dr. Eng. Ahmed Ratnani is the leader of Magnetohydrodynamic MHD group at Max-Planck IPP, Munich, Germany. He hold his Phd. from University of Starsbourg, INRIA, France, a Master degree in Applied Mathematics from University of Paris VI, France and an engineering diploma from ENISMAG, France, with a major in Mathematical Finance.
His research topic covers Partial Differential Equations, their discretization with FInite Elements, Applied Linear Algebra, Language Theory and High Performance Computing.
Among the different features of Python, Functional programing (FP) is gaining, nowadays, more and more interest. The Python syntax makes it simpler and intuitive. From the theoretical point of view, FP finds its roots in Lambda Calculus, a mathematical theory developed by Alonzo Chruch in 1930s to study computations with functions. On the other hand, the last achievements in the Haskell language and the aim to ensure portability on different architectures put FP on the front scene once more.
In this talk, we will revisit the foundation of FP and show how one can build a multi-architecture compiler for Python, for some specific parallel patterns such as the Map-Reduce.