Supercomputing Academy: Parallel Programming with MPI
Learn to write parallel programs using MPI — the industry-standard protocol for distributed memory systems. This flexible online course by HLRS takes you from the basics to advanced techniques, preparing you to run efficient simulations on clusters and supercomputers.
Praktische info:
Inschrijven?
- Voorwaarden: Basic knowledge in C or Fortran or Python, basic knowledge in Linux, basic understanding of computer hardware
- Prijs: €0 - €1370
Leertraject
MPI is a communication protocol for parallel programming based on message exchange between individual processes. These processes can be executed on distributed memory systems with multiple nodes. This makes MPI scalable on systems larger than single computers.
MPI offers a range of tools to maintain the flow of information between individual processes, which enables the execution of a parallel program that can be divided into several smaller parts. In the course of necessary communication, overhead always occurs, which normally limits the scalability of a parallel program. However, a properly optimized program opens up the possibility of using MPI on a distributed memory system (e.g., cluster or supercomputer) with satisfactory efficiency, where thousands or tens of thousands of nodes can be used.
This flexible online course offers the opportunity for intensive exchange with the instructors and among the course participants.
Agenda
- Week 1: The Beginners' MPI
Introduce the fundamentals of MPI, including the process model, basic communication techniques, and non blocking communication. - Week 2: Beginner & Intermediate MPI
Dive deeper into intermediate MPI concepts, including collective communication, error handling, derived datatypes, and virtual topologies. - Week 3: Advanced MPI - Part 1
Explore advanced MPI topics, including one-sided communication, shared memory communication, and synchronization rules. - Week 4: Advanced MPI - Part 2
Continue with advanced MPI topics, focusing on collective communications, and virtual topologies.
It wraps up with the parallelization of a simulation of the 2-dimensional heat equation. - Week 5: Advanced MPI - Additional optional content
Parallel file I/O, advanced topics on communicators and derived datatypes, and a short tour through other MPI topics. It wraps up with a lecture on best practices.
- Seminars are scheduled on Mondays, 17:30-19:00: Nov. 2 (kick-off), and Nov. 9, 16, 23, 30 (discuss the content of weeks 1-4). It is highly recommended to participate in these seminars, but if you have a conflict, you may view the recording.
Gerelateerde opleidingen
Introduction to Simulation Based Inference: Enhancing Synthetic Models with Artificial Intelligence
Tutorial - Online - Forschungszentrum Jülich
Supercomputing Academy: Deployable Data Analysis & AI Pipelines with HPC
Opleiding - Online - HLRS