This image appears in the ARCHER2 Training Courses to help guide learners through the available materials. It provides a helpful overview of what's available, and roughly delineates what resources would be most useful to researchers coming from various backgrounds (or focus areas). Recycling the concept, if not the image itself, would be a boon to HPC Carpentries learners.
Learn More →
Learn More →
Moderator(s)Alan O'Cais — University of BarcelonaMarc-André Hermanns — RWTH Aachen UniversityAbstract"HPC Carpentry is the first entry to the Carpentries Lesson Program Incubator. As such it has a proposed core curriculum, which will be outlined at the beginning of this session. HPC Carpentry exists in an ecosystem that is constantly developing however. Beyond the core curriculum, we would like to discuss other existing and proposed lessons that are relevant to HPC Carpentry and how to adequately integrate them so that they complement the core and each component is sufficiently maintained. Questions we would like to address are:
Nov 7, 2024Updated 7 October 2021 A much better starting point is here: HPC-in-a-day Pre-workshop Survey Survey Overview HPC Carpentry is an unofficial, Carpentries-style suite of materials intended to ease the adoption of high-performance computing resources. We welcome feedback of all kinds, and look forward to seeing you on our screens. Introduction and Purpose Thank you for volunteering to take part in our research study, which is about understanding the effectiveness of our workshops. To participate in the study, you will complete a short survey about your skills and perspectives related to our workshop content before and after your workshop. Depending on your location, the survey will be 16-18 questions long and will take approximately 5-10 minutes to complete.
Oct 7, 2021An introduction to parallelisation and workflow automation by Alan O'Cais and Peter Steinbach With this single day introduction, we want to take your HPC cluster skills to the next level. We plan to introduce automated pipelines and parallelization suited to our learners. We assume that learners are able to submit single jobs to a SLURM based scheduler and have a basic understanding of the UNIX shell and Python. For parallelization, we aim to provide a thorough introduction on how to approach implementing data parallelism in Python. For this, we will use shared memory parallelisation using multiprocessing and distributed memory parallelisation using message-passing-interface for a compute intensive problem. The day will be concluded with an introduction on how to automate pipelines on a cluster - which are typically found with data intensive tasks. All teaching will be performed hands-on on a custom cluster provided to the students. Some House keeping :handshake: please be advised that we conduct this lesson under The Carpentries Code of Conduct :pencil: this is the shared document for collaboarative notes, please use it for notes, links, code snippets that support the content Agenda
Mar 24, 2021or
By clicking below, you agree to our terms of service.
New to HackMD? Sign up