Why invest the time and effort to learn HPC?
Learning how to use HPC will likely not be valuable for all users in all cases, but it can be exceptionally valuable and save significant time and effort in certain situations. Learning these skills provide HPC knowledgable users with a strategic advantage that can serve them throughout their academic, research, and professional careers.
Anyone with a workstation crashing due to memory problems from too much data, running time consuming computational tasks, performing repetitive computations, wanting to scale their computational work, or just needing more compute power should consider using HPC resources for improved speed, efficiency and time-savings.
Common HPC Skills
Becoming an expert in HPC requires knowledge in Linux Command Line, and mastering a number of tools. The tools that provide the greatest value to you may be different from those in other scientific domains. While it may seem initially daunting, UW ARCC helps all researchers regardless of HPC experience level. Our department has developed a number of tutorials and we offer bi-weekly office hours allowing you to learn HPC at your own pace. Below is a very general list of common set of skill to master to become an expert HPC user:
General HPC Use and Knowledge
The Linux Operating System
Command Line Interface (CLI)
Data Transfers and Management
Common Software for HPC Clusters (Conda, Python, R, etc)
Scheduled Trainings
Our bootcamp materials can help train new or intermediate users on ARCC HPC resources, and will cover the topics in modules listed in the table below:Schedule | Topics/Modules (Expand for Description) |
|---|
Monday
Beginning at 9am
ITC 259 | Linux Command Line: The File System Content Description Introduce users to typing commands using the command line to work with the Linux operating system. Focusing on hands-on exercises, it will introduce the basic structure and use of the file system, and how to find help. Linux Command Line: View, Find and Search Content Description Introduce how to work with, view, find and search the content of text files. The workshop is aimed at beginners with basic command-line experience of the Linux file system and will focus on hands-on exercises. Prerequisites: Basic command-line experience or attendance at Intro to Linux Command Line: The File System Text Editors in Linux Content Description Introduce the various text editors available on our Services. Using emacs as a GUI via an OnDemand interactive desktop, and vim/nano from a terminal command line. Prerequisites: Basic command-line experience or attendance at either Intro to Command Line module |
Tuesday
Beginning at 9am
ITC 259 | Intro to HPCs Content Description Introduce UW ARCC, our core services, and how we can help.
Introduce the core concepts behind High Performance Computing and clusters.
Introduce available HPC Resources to UW Researchers Intro to Cluster Access Content Description Introduce users on how to remotely access the clusters Over the web using OnDemand Accessing Beartooth with traditional SSH and 2 factor authentication Accessing MedicineBow using passwordless authentication
Optional IT Data Center Tour Over Lunch Hour Intro to LMOD: The Module System Content Description Introduce users to LMOD, ARCC’s HPC software module system.
Introduce how to set up and personalize their environment and software for their computational needs. Intro to Job Scheduling Content Description Introduce users to job management using the Slurm system - demonstrating how to create interactive jobs and submit jobs to the cluster queue that follow a basic workflow. |
Wednesday
Beginning at 9am
ITC 259 | Using OnDemand (ARCC HPCs in a web browser) Content Description Introduce OnDemand Show users how to navigate remote web access to HPC Services. Conda: Creating Python/R and other Application Environments Content Description Introduce what Conda is and how it can be used to create a variety of environments, such as Python/R and contained applications. Review: Consolidating what we covered so far |
Thursday
Beginning at 9am
ITC 259 | Data Management Data Management Module Description Introducing the concept of data management with a focus on some common best practices of working with data on ARCC systems. Data Transfers Content Description The practical options for transferring data from onto HPC systems from local computers and remote servers. Using Python, Conda and Pip on the Cluster Content Description Discuss how the various Python related tools work together on the cluster and introduce a series of best practices for managing these environments. (**Note: This is not an introduction to programming in Python. It covers how to use Python, Conda and Pip on an HPC cluster) Prerequisites: Intermediate users of HPC or attendance at prior modules, and experience programming in Python Using R and R Studio on the Cluster Content Description Discuss how the various R related tools and RStudio work together on the cluster and introduce a series of best practices for managing these environments. (**Note: This is not an introduction to programming in R. It covers how to use R and RStudio on an HPC cluster) Prerequisites: Intermediate users of HPC or attendance at prior modules, and experience programming in R or RStudio |
Friday AM
Beginning at 9am
ITC 259 | Jupyter Notebooks with OnDemand Content Description Demonstrate how to use the OnDemand Jupyter Service across a variety of available languages and kernels. Demonstrate how to convert an existing Conda environment into a kernel that can be used within a notebook.
Prerequisites: Intermediate users of HPC or attendance at Using Python, Conda and Pip on the Cluster Put it into Practice: Wrap up and ARCC Q&A Content Description If you have any questions or need help with something that wasn’t covered in the bootcamp, we have allocated some of the morning to go over concepts that were not covered, or things you may want additional help with. |
Those who sign up for a bootcamp should plan to attend the entire week of topics.
Please contact arcc-help@uwyo.edu with any questions, inquiries, or comments.