WVU Research Computing Workshop Series
Introduction to High-Performance Computing
Online Only / Zoom Link: https://wvu.zoom.us/j/96313400972
FOR: Aimed at new users of WVU’s HPC Clusters. Intended mainly for faculty, and graduate students across West Virginia whose research problems cannot typically be solved using local computing resources such as desktop computers or laptops.
PURPOSE: To give the attendees an overview of topics in High Performance Computing, parallel computing, machine learning, and data science. A more specific purpose is to provide attendees with practical skills that can be applied quickly in HPC to take advantage of the computational resources offered by WVU.
TIME COMMITMENT:
- August 18 (Monday) 10AM - 12PM
- August 21(Thursday) 10AM - 12PM
- August 25 (Monday) 10AM - 12PM
- August 28 (Thursday) 10AM - 12PM
Each lecture is divided into four micro-learning sessions, designed to be 20 minutes long with a single learning objective and exercises to reinforce the learning concepts and fix the skills. Each session will invite attendees to write and submit jobs to the cluster as part of the exercises. The focus is not on learning programming but on using the tools needed to get the results using existing codes. The examples span various research areas, ranging from physics to biology, and from pure mathematics to data science.
To offer content that benefits a broad audience, attendees are not required to have familiarity with the UNIX/Linux shell or previous experience in HPC. The workshop starts from the basics and builds a broad but effective set of skills and outlines a map of the HPC landscape opening doors to new venues and more advanced workshops in the future. Topics such as parallel programming, machine learning, including AI and data science are kept general enlightening pathways for future learning.
OUTLINE OF TOPICS:
1. HPC Concepts: An overview of Hardware and Software
2. The lifecycle of a HPC job: Submitting a first job
3. The UNIX/Linux Command Line Interface
4. Terminal based text editors
5. The Workload manager: Slurm
6. Terminal multiplexer: tmux
7. Accessing Software: Environment modules, Conda and Singularity
8. Programming the Shell: Bash scripting
9. Scientific Programming languages (Fortran, C/C++, Python, R and Julia)
10. Parallel computing (OpenMP, OpenACC, MPI and CUDA)
The High Performance Computing (HPC) group, a unit within the WVU Research Office, provides hardware and software resources, training, and strategic guidance to support computationally intensive research, serving the needs of the WVU research community.