Introduction to High Performance Computing

This course aims to introduce the principles of high performance computing (HPC) systems, how they operate, and how we can utilise the computing power they can offer to model complex systems.

This will start with a discussion of supercomputers (HPC systems), their purpose, hardware, and trends in computing. We will then cover material on the fundamentals of modern computers and how we can perform computations in parallel. This will lead into a discussion of how supercomputers derive their performance, and how distributed memory architecture is used in HPC systems.

We will run through a concrete simulation example, discussing the details of the problem and how it can be modelled on a computer. We will then detail how parallel programming can be applied to problems to speed up a calculation, before discussing performance in parallel programs. Finally we will discuss how simulations are developed and used to make predictions, and some real world examples.

Prerequisites

To take part in the exercises you will need:

  • Access to the ARCHER2 system

  • Basic Knowledge of Git

  • Basic Knowledge of Bash


Arrangements for this self-service course

This course has already run in other forms in the past and we are keen the material remains available to the community, therefore the course has been presented in this new format with support from the EuroCC project. This course is hosted on github and will run in an unfacilitated form meaning the course will not have involvement, input and direction from the Educators. However it will be monitored, so please therefore raise issues on the git repository which the course is hosted from, if there are comments on the material.

Questionnaires

Before you begin this self-service course, we would be most grateful if you could complete this short pre-course questionnaire.

Once you have completed the course, please complete the post-course questionnaire.

Feedback

When you have finished the course, please fill in the Feedback Form. This is designed to be anonymous, although you can leave your contact details if you wish. We use the same form for all training courses, so some questions (e.g. regarding the venue) may not be relevant.


Welcome

This may be the first self-service course you have worked through, so to help you get the best learning experience we have prepared a few tips for you. Hopefully, they will help you get the most out of the course.

The course material is presented as a mixture of videos, articles, exercises and quizzes. The course is no longer fully facilitated so we would like to encourage you to interact with others taking the course if many of your local team are working through the content. If there is anything that you find confusing, ask, and if you know an answer to a question, answer it. You will find that both asking and answering questions helps to consolidate not only your but also other people’s knowledge.


../../_images/BayesInterior.jpg

About EPCC

Since 1990, EPCC has developed a reputation for leading-edge capability in all aspects of High Performance Computing (HPC). Based at the University of Edinburgh, we work with industry, academia, public and third-sector organisations to promote the adoption and value of HPC.

With a team of more than 110 highly qualified specialists, we have an exceptional pool of talent. Our engineers and technical staff have a powerful combination of practical and theoretical knowledge that keeps EPCC at the forefront of industry and research.

EPCC aims to accelerate the effective exploitation of novel computing throughout industry, academia and commerce.

This has been our mission for over 30 years as we strive to be the UK’s premier Supercomputing centre, renowned internationally for innovation and leading-edge research.

Our work is guided by key goals and objectives that support this mission and place EPCC at the heart of the Exascale era.

Delivering world-class services We will continue to deliver world-class High Performance Computing (HPC) and data services to benefit our users and partners. As well as supporting businesses with our advanced computing facilities and capabilities.

Leading data-driven innovation We are a pivotal element in the success of the Edinburgh and South East Scotland City Region Deal and its Data-Driven Innovation initiative.

Collaborating globally We will continue to build and maintain collaborations and strong relationships with key organisations worldwide, undertaking globally-important computing research.

Promoting teaching excellence We endeavour to maintain our position as a major provider of HPC training.


../../_images/large_hero_a8f32791-5354-4a66-aeb0-79352dedae18.jpg

This course has been adapted with EuroCC funding to update the original course developed under PRACE and improve course flexibility to allow it to be tailored to multiple machines’ needs while keeping the core theme of the course.

Acknowledgements

This material has been developed by James Richings, Stephen Farr, Manos Farsarakis, David Henty and Weronika Filinger at EPCC under funding from EuroCC. This was developed from original material developed, with additional contributions from Zheng Meyer-Zhao at Surfsara, under funding from PRACE.

This new course benefits from the Sphinx template developed by the Swedish NCC which was also funded by EuroCC.