[FreeFEM 10] Parallel programming and Message Passing Interface (MPI)

Parallel programming and MPI are crucial tools for achieving high performance computing. With parallel programming, a program can be broken down into smaller tasks that can be executed simultaneously on multiple processors, leading to faster execution times. MPI (Message Passing Interface) is a widely used standard for exchanging messages between multiple processors in a parallel computing environment. By using MPI, programmers can write code that runs on a cluster of computers, distributing the workload across them for efficient execution. In this video, we explore the basics of parallel programming and MPI and demonstrate how they can be used to speed up computations. Since FreeFEM provides an easy-to-use parallel programming features, we use it for this demonstration.

Codes, models, and resources:
You can find all the codes and models required to follow the videos and reproduce the output, grouped for different episodes, at http://tuxriders.com/videos/freefem/ and https://github.com/TuxRiders/freefem-...

💡 Finite element series, in which a single problem was solved in different open-source tools including FreeFEM    • Introduction to finite element modeling  

💡 Applied numerical computing series, in which the underlying theories of these videos are discussed:    • Introduction to applied numerical modeling  

Topics covered:
🎯 The necessity of Message Passing Interface in scientific computing
🎯 Concept of MPI size and MPI rank
🎯 Transferring data among computing nodes
🎯 Broadcasting and collective operations
🎯 Concept of distributed memory parallelism
🎯 Parallel programming terminologies

Lecturer: Mojtaba Barzegari https://mbarzegary.github.io/

To learn more about the goals of the TuxRiders project, please visit our website at http://tuxriders.com.

Chapters in this video!
################
00:00 - Intro
01:05 - What is MPI?
05:58 - MPI size and MPI rank
10:01 - Concept of MPI barriers
15:21 - Controlling the order of execution
18:20 - Broadcasting data and collective operations
24:56 - Sending/receiving data via MPI requests
28:53 - MPI reduction to collect data
31:56 - See it in action: compute the value of Pi in parallel