COURSE OBJECTIVES

• Introduce HPC on distributed architecture’s fundamental principles and good practices.    
• Master the use of MPI library. 
• Practical application with well-known examples (research and industry). 

Engineers and researchers aiming to use HPC and gain initial experience.  

Presentation of High-Performance Computing 
• Main issues and necessity of parallelization. 
• Examples of applications.
• Hardware and software components of High-Performance Computing (processors, memory, applications allowing the implementation of parallelization on the hardware).
• Measurement of computational performance and improvement techniques. 
• Introduction to parallelism performance indicators.

Architecture of HPC systems 
• Shared memory architecture. 
• Distributed memory architecture.
• Hybrid architectures.

High-performance architecture management systems: Scheduling and Load-balancing with SGE 

Implementation of parallelization 
• Concepts of message exchange. Programming interfaces. 
• Parallel computing in MPI (Message Passing Interface).
– MPI-1: Number of processes, process number. Point to point communications. Collective communications. Communicators. Topologies.
– MPI-2: Dynamic process management. Parallel I/O. Memory to memory communication.
– Libraries and interfaces.

Practical work 
• Getting started with MPI in Python through MPI4Py.
• Parallel simulations in MPI. 
• Implementation of optimization algorithms with MPI.