When dealing with performance problems of programs involving a large amount of numerical operations, an HPC programmer has to know the existence of specifically developed numerical libraries, and how to use them. Their high scalability and portability can be of great help, since they can save a good amount of computing time and memory, and avoid the effort of having to think and implement functions already covered by one of such libraries.

In this course, some of the most important numerical libraries for HPC will be presented: In particular, there will be covered libraries for Linear Algebra (Scalapack and others), Fast Fourier Transform (Fftw) and modelling with Partial Differential Equations (PetsC). Time for practicing with these libraries is also scheduled.Topics

    Introduction to HPC Numerical libraries.

    Introduction on Linear Algebra libraries, (Parallel) Basic Linear Algebra Subprograms (PBLAS), (Scalable) Linea Algebra PACKage (ScaLAPACK), Basic Linear Algebra Communication Subprograms (BLACS), PBLAS and ScaLAPACK, BLACS + MPI, Outline on Matrix Algebra for GPU and Multicore Architecture (MAGMA) library.

    Introduction on Fast-Fourier transforms theory, Parallel Data Distribution (Slab and 2D decompisition), FFTW library (instructions, examples, exercises), Introduction to 2DECOMP&FFT libraries.

    Introduction to PetsC, PetsC and Matlab, introduction to TAO.

Prerequisites
    Knowledge of a scientific programming language (C/C++, Fortran) , basic notions of parallel programming (MPI, recommended openMP)Price: free of charge COURSE AGENDA

https://events.prace-ri.eu/event/125/