Julia for high-performance scientific computing
Julia is a scientific programming language that is free and open source - see https://julialang.org/ for downloads, documentation, learning resources etc. Bridging high-level interpreted and low-level compiled languages, it offers high performance (comparable to C and Fortran) without sacrificing simplicity and programming productivity (like in Python or R).
Julia has a rich ecosystem of libraries aimed towards scientific computing and a powerful in-built package manager to install and manage their dependencies. Julia is also gaining ground in HPC as it supports both threading and distributed-memory parallelisation as well as GPU computing.
After motivating why Julia can be a good option for your next HPC project, this lesson goes through various approaches for writing performant and parallel code, including how to write fast serial code, how to write multithreaded and distributed code, how to run Julia on an HPC system, and how to port Julia code to GPUs.
If you are new to the Julia language, please make sure to go through this introductory Julia lesson before going through this HPC lesson independently or attending a workshop where it is taught.
To learn about using Julia for problems in data science, please visit the lesson Julia for high-performance data analysis.
Experience in one or more programming languages.
Understanding of basic Julia syntax, best practices and development tools, corresponding to what is covered in the ENCCS Julia-intro lesson
Familiarity with basic concepts in high-performance computing (HPC) (including parallelization by threading or multiprocessing) and programming graphical processing units (GPUs). No direct experience is required.
Who is the course for?
This lesson material is targeted towards students, researchers and developers who:
are already familiar with one or more programming languages (Python, R, C/C++, Fortran, Matlab, …)
want to add a new exciting high-level yet performant language to their repertoire
might be mixing a high-level and a low-level language for performance reasons but want to make their life easier
need to analyze big data or perform computationally demanding modeling, analysis or simulations
want to develop high-performance parallelized and/or GPU code but prefer to stay within a productive high-level language.
About the course
This lesson material is developed by the EuroCC National Competence Center Sweden (ENCCS) and taught in ENCCS workshops. It is aimed at researchers and developers who want to learn a modern, high-level, high-performance programming language suitable for scientific computing, data science, machine learning and high-performance computing on CPUs or GPUs. Each lesson episode has clearly defined questions that will be addressed and includes multiple exercises along with solutions, and is therefore also useful for self-learning. The lesson material is licensed under CC-BY-4.0 and can be reused in any form (with appropriate credit) in other courses and workshops. Instructors who wish to teach this lesson can refer to the Instructor’s guide for practical advice.
Graphical and text conventions
Different graphical elements are used to organize the material.
Type-along sections are intended for live coding where all participants type-along and appear in a separate text box, marked with a keyboard emoji:
Defining a variables
This is how you set a variable in Julia:
x = 1
All lesson episodes (sections) end with one or more exercises for participants to practice what they’ve learned, marked with a hand-writing emoji. Sometimes there’s also a solution:
Printing to screen
Which of these commands prints the value of the variable
Correct answer is both 1 and 2!
print() and adds a new line.
Sometimes important information is displayed inside boxes with an exclamation mark emoji:
Please don’t hesitate to ask questions during the workshop!
Discussion exercises are conducted either via voice or through a shared workshop document.
Are these instructions clear?
Discuss any questions about the lesson format either via the shared workshop document or in breakout room sessions.
Many resources for learning Julia can be found at https://julialang.org/learning/. The list includes Julia Academy courses, the Julia manual, the Julia Youtube channel, and an assortment of tutorials and books.
A recent talk given by Kristoffer Carlsson, developer at Julia Computing in Sweden, gives an excellent overview on using Julia for HPC.
This repository is from the Julia for High Performance Computing Course @ HLRS developed by Carsten Bauer.
Several examples and formulations are inspired by other Julia lessons, particularly:
Introduction to Julia provided by CSC and Aalto
All ENCCS instructional material is made available under the Creative Commons Attribution license (CC-BY-4.0). The following is a human-readable summary of (and not a substitute for) the full legal text of the CC-BY-4.0 license. You are free:
to share - copy and redistribute the material in any medium or format
to adapt - remix, transform, and build upon the material for any purpose, even commercially.
The licensor cannot revoke these freedoms as long as you follow these license terms:
Attribution - You must give appropriate credit (mentioning that your work is derived from work that is Copyright (c) ENCCS and, where practical, linking to https://enccs.se), provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
No additional restrictions - You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits. With the understanding that:
You do not have to comply with the license for elements of the material in the public domain or where your use is permitted by an applicable exception or limitation.
No warranties are given. The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material.