This tutorial is about getting in touch with the Julia programming language that "makes it easy to express many object-oriented and functional programming patterns". It mainly focuses on (i) set up the Julia environment, (ii) run a set of simple examples on creating matrices, plotting charts, and executing simple for-loops with CUDA, and (iii) go through introductory examples on machine learning (Regression and Decision Trees).
The MAGMA project aims to develop a dense linear algebra library similar to LAPACK but for heterogeneous/hybrid architectures, starting with current "Multicore+GPU" systems.
ScaLAPACK is a software package provided by Univ. of Tennessee; Univ. of California, Berkeley; Univ. of Colorado Denver; and NAG Ltd.
OpenBLAS is an optimized BLAS library based on GotoBLAS2 1.13 BSD version.
MPICH is a high performance and widely portable implementation of the Message Passing Interface (MPI) standard.
The Intel® oneAPI DPC++/C++Compiler provides optimizations that help your applications to run faster on Intel® 64 and IA-32 (Windows and Linux only) architectures, with support for the latest C, C++, and DPC++ language standards (including C++17).
https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compiler.html
The AMD Optimizing C/C++ and Fortran Compilers (“AOCC”) are a set of production compilers optimized for software performance when running on AMD host processors using the AMD “Zen” core architecture.
The NVIDIA HPC Software Development Kit (SDK) includes the proven compilers, libraries and software tools essential to maximizing developer productivity and the performance and portability of HPC applications.
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
ARPACK software is capable of solving large scale symmetric, nonsymmetric, and generalized eigenproblems from significant application areas. The software is designed to compute a few (k) eigenvalues with user specified features such as those of largest real part or largest magnitude. Storage requirements are on the order of n*k locations. No auxiliary storage is required. A set of Schur basis vectors for the desired k-dimensional eigen-space is computed which is numerically orthogonal to working precision. Numerically accurate eigenvectors are available on request.