Running Matlab on HPC
The Matlab course offers an opportunity to learn the application of the MATLAB programming language on the HPC cluster – parallel computation issues are considered to allow the computational task to be divided into different computing cores or computing processors, both in terms of computing power and memory.
This course does not offer the basic knowledge of Matlab, however, covers specific Matlab tools for the use of HPC or supercomputer. It is advised to have basic knowledge of Matlab, such as “Introduction to Matlab“.
Lecturer: Dr.Sc.Ing. Maris Terauds.
MATLAB is a high-level language and interactive environment for numerical computation, visualization, and programming.
The aim and tasks of the course
The aim of the course is to develop skills in MATLAB program for performing parallelization from both the calculation and memory point of view.
Tasks:
- To introduce course participants to the functions and capabilities of the parallel computing toolbox;
- View the nuances of parfor use;
- View the nuances of using the pareval and spmd commands;
- View the nuances of dividing an array between different workers, and nodes.
The training courses will take the form of interactive (participatory) lectures, repeating what is said on your computer and asking questions.
Topics
The following topics will be covered:
- Short overview of program acceleration techniques that are not closely related to parallel computing. HPC (High-Performance Computing) nodes. Connecting to HPC.
- Short overview of parallel computing toolbox functions and possibilities.
- Parallel execution of the program using parfor cycles and their application nuances.
- Parallel execution of the program using spmd and pmode loops and their application nuances.
- Short overview of arrays, Big data problem, etc.
- Partitioning arrays and matrices between multiple cores or cluster workers.
