Training

OpenMP with AMD GPU Hackathon

3rd-5th April 2023

Introduction

This is a hands-on hackathon where participants will apply presented techniques to their own codes.

The workshop & pre=event training will cover:

20th-22rd March pre-event online training:

  • Getting Started with OpenMP & offloading
  • AMD Communication Fabrics and GPU-Aware MPI
  • Introduction to Rocprof

3rd-5th April in-person hackathon at UCL with AMD & DiRAC support.

Format

This will be an in-person hackathon, held at UCL. Participants will use AMD’s dedicated development cluster. This will be a pre-event training session on the 20th-22nd of March. This training will be held online.

The workshop will run over 3 days. Each day will normally be focused on a practical session where participants have an opportunity to apply what was taught in the pre-event training to their code. Support will be there from AMD and from DiRAC support teams. During the event, there will be feedback and Q&A sessions to help spread good practice and address any issues.

In the afternoon of the 3-days, teams will be given time to develop a presentation that will be presented by themselves to all participants. It is intended that this will be recorded to help other DiRAC users.

Target Audience

The target audience are researchers who want to:

  • Optimising their code for AMD’s latest GPUs.
  • Run their code efficiently to get the best performance.
  • Learn about and take advantage of the new AMD features and tools.
  • Build links with other research groups and AMDs technical team.

Requirements

At least one member of a team needs good experience in a programming language. Also in-depth knowledge of your own codes. Technical support can be arranged if required.

If you have any questions please contact richard.regan@durham.ac.uk

Registration

Registration is closed

By Richard Regan Admin, ago

Performance Analysis Workshop Series 2023

Performance Analysis Workshop Series 2023
20th April – 18th May 2023

With every new generation of computers, we see the gap between the theoretical performance of a machine and the performance that is actually delivered by applications widen. Codes struggle to exploit the hardware. It has therefore become critical that researchers and research software engineers in HPC to understand how well and why codes use the machinery as they do. Insight into performance behaviour can drive the code evolution and ultimately become the means through which future advancement through computing are facilitated.

This workshop series offers a comprehensive introduction to a selection of open source tools that enable researchers to assess the performance behaviour of their code. The workshops will be augmented by revision sessions of some of the core HPC know-how. We encourage participants to bring along their own codes so they can continually assess and improve them throughout the series.

For more information and to register, click here.

By jamblay, ago

DPU Hackathon 2023

DPU Hackathon 2023

16/17 February 2023
Durham University, Department of Computer Science, Durham, UK (hybrid, in person preferred)
In collaboration with NVIDIA Networking

Durham’s Department of Computer Science, in collaboration with Durham’s DiRAC facilities and Durham’s ExCALIBUR H&ES installations, has organised a 1.5 day hackathon on how to use NVIDIA BlueField technology.

BlueField-empowered systems are supercomputers, where each individual networking card is equipped with additional ARM processors. These processors can, for example, take ownership of data movements between nodes, i.e. release the host from messaging-related work, manipulate message content while the messages fly through the network, own checkpointing,…

During the hybrid workshop, participants will first get a brief intro into BlueField technology, and can then try out prepared exercises on these machines. After that, we host a series of talks and brainstorming sessions on how this technology could enable next-generation simulation software. Finally, NVIDIA’s experts will be available to help with some prototyping of ideas on BlueField cards.

For more information and to register, click here.

By jamblay, ago

AMD Tools Workshop

28th-30th November 2022

Introduction

This is a hands-on workshop where participants will apply presented techniques to their own codes.

The workshop will cover:

  • Optimal Pinning processes for the AMD architecture
  • AMD Compilers and libraries
  • μProf tutorial

Format

This will be an in-person workshop, held in Durham. Participants will use their normal DiRAC site.

The workshop will run over 3 days. Each day will normally start introducing a topic which is then followed by a lengthy practical session where participants have an opportunity to apply what was taught to their code. Support will be there from AMD and from DiRAC support teams. During the event, there will be feedback and Q&A sessions to help spread good practice and address any issues.

On the afternoon of the 3-days, teams will be given time to develop a presentation that will be presented by themselves at DiRAC day on the 8th of December at UCL.

Target Audience

The target audience are researchers who want to:

  • Optimising their code for today and tomorrow on AMD CPUs.
  • Get the most out of our new DiRAC-3 AMD systems.
  • Run their code efficiently to get the best performance.
  • Learn about and take advantage of the new AMD features and tools.
  • Build links with other research groups and AMDs technical team.

Requirements

At least one member of a team needs good experience in a programming language. Also in-depth knowledge of your own codes. Technical support can be arranged if required.

Registration

Registration closed.

By Richard Regan Admin, ago

Quantum Workshop

Please enter your answer for the quantum script below.

Quantum Answers

By Richard Regan Admin, ago

April 2022 Machine Learning for Science Training Course Session Recordings

Dates: 4th April to 8th April 2022

The following links are only available to the attendees of the April 2022 training course. For details of upcoming courses, please follow the DiRAC twitter page.

Please follow the links below to view the recordings of the course lectures.

Classical Machine Learning

Introduction to Machine Learning

Supervised Learning Tequniques

Unsupervised Learning Techniques

Deep Neural Networks

Deep Neural Networks

Convolutional Neural Networks & Pretrained Models

Image Processing

Long Short-Term Memory (LSTM) Networks

Back Propagation and Gradient Descent

Debugging and Exploring ML Solutions

Generative Models

Generative Models Introduction

Generative Models Autoencoders

Generative Models VAEs

Generative Models GANs

Helping you prepare for the future

By jamblay, ago

AMD Induction Training (AIT)

29th September 2022

Sorry registration is closed.

Introduction

This is a hands-on workshop where participants will use code snippets to illustrate the best approach to accessing the power of our new AMD ROME CPUs based at Durham & Leicester. This training will also be useful to anyone using the AMD CPUs on the GPU systems at Cambridge and Edinburgh.

The workshop will cover:

  • The ROME Microarchitecture & Memory channels
  • Pinning processes
  • AMD Compilers and libraries
  • MPI considerations
  • uProf tutorial
  • Hands-on examples spread throughout the day

Format

The virtual workshop will be run on multiple DiRAC sites so you will be able to experience exactly what you need to do on the system you use.

  • Support will be there from AMD and the local technical support teams.
  • A shared slack channel will be available to ask questions, highlight any issues and share good practice between sites.

Target Audience

The target audience is researchers:

  • Those who are interested in building and running their code on our new DiRAC-3 AMD systems.
  • Those who want to run their code efficiently to get the best performance.
  • Those who want to take advantage of the new AMD features and tools.

Requirements

Only basic experience of a language is required C/C++, or Python.

Registration

Sorry registration is closed.

By Alastair Williams, ago

Programme

Dates: 4th April to 8th April 2022, Every Day 09.00 – 13.00

Day 1: Classical Machine Learning

09:00 – 09:30Introduction to Machine Learning
09:30 – 10:15Lecture – Supervised Learning Techniques
10:15 – 11:15Hands-On Practical
11:15 – 11:45Break
11:45 – 13:00Lecture – Unsupervised Learning Techniques

Day 2: Deep Neural Networks

09:00 – 09:45Lecture – Neural Networks
09:45 – 10:45Hands-On Practical
10:45 – 11:15 Break
11:15 – 12:00Neural Networks (Backprop)
12:00 – 13:00 Hands-On Practical

Day 3: Image Processing

09:00 – 09:45Lecture – Convoluted Neural Networks
09:45 – 10:45 Hands-On Practical
10:45 – 11:15  Break
11:15 – 12:00 Lecture – Autoencoder
12:00 – 13:00 Hands-On Practical

Day 4: Generative Models

09:00 – 09:45 Introduction to Generative Models & VAE
09:45 – 10:30 Hands-On Practical
10:30 – 11:00Break
11:00 – 11:45 Lecture – Generative Adversarial Networks (GANs)
11:45 – 12:30Hands-On Practical
12:30 – 13:00 Lecture – Debugging/ Exploring

Return to Course Overview

Helping you prepare for the future

By Alastair Williams, ago

Machine Learning Techniques for Science

4th April 2022 to 8th April 2022

Course: Machine Learning Techniques for Science

Science is undergoing a data explosion and the advent of Artificial Intelligence and Machine Learning techniques is revolutionizing the way scientists tackle their research. Simulations and observations now generate Petabytes of data and machine learning is providing novel and powerful methods for analysing those experimental datasets and extracting essential science in ways that have not been possible before.

Description

The course provided a practical, and hands-on introduction to the concepts, methods, and toolkits for applying machine learning to fundamental scientific problems. Held virtually over five consecutive mornings, lectures, including worked examples to reinforce the concepts, and hands-on practical sessions were delivered. The practical sessions, where participants had access to dedicated GPU resources, were meant to ensure that participants learned to apply their newly learned skills on practical problems stemming from physics, astronomy and other domains. The course covered.

  • Conventional machine learning techniques (such as decision trees),
  • Neural and deep neural networks, the cornerstone of modern AI
  • Generative models to enable you to generate synthetic, yet realistic, datasets with labels, and
  • Debugging and RSE aspects of the machine learning

The course was provided by DiRAC in collaboration with STFC’s Scientific Machine Learning Research Group (SciML) in the Scientific Computing Department.

An Outline Timetable can be found here

Prerequisites

Basic programming skills in Python.

Previous Course Testimonials

“I found the course to be the ideal mix between theory and practical examples. Very informative on how to use ML for cutting-edge science research questions.”

“This course was the perfect introduction to machine learning from a scientific perspective; instructors were enthusiastic and very helpful, and the lectures and materials were approachable for someone with no previous experience, while remaining detailed enough to be of practical use. I felt I gained a good overview of a range of important techniques and algorithms and will certainly apply the knowledge gained during my chemistry PhD.”

Registration

Places on the Machine Learning Techniques for Science course are allocated giving priority to individuals who are part of a DiRAC project. If you would like find out how to become a DiRAC user, please see our latest call for proposals.

Registration is now closed.

Follow us on Twitter @DiRAC_HPC for future ML course announcements.

Helping you prepare for the future

By Alastair Williams, ago