These resources have been contributed and “vetted” by the community of cyberinfrastructure professionals (researchers, research computing facilitators, research software engineers and HPC system administrators) that are participating in programs such as this one, that are supported by the ConnectCI community management platform. Additional Knowledge Base Resources are always welcome!
A comprehensive list of training resources from the HPC University. HPCU is a virtual organization whose primary goal is to provide a cohesive, persistent, and sustainable on-line environment to share educational and training materials for a continuum of high performance computing environments that span desktop computing capabilities to the highest-end of computing facilities offered by HPC centers.
Cornell Virtual Workshop is a comprehensive training resource for high performance computing topics. The Cornell University Center for Advanced Computing (CAC) is a leader in the development and deployment of Web-based training programs. Our Cornell Virtual Workshop learning platform is designed to enhance the computational science skills of researchers, accelerate the adoption of new and emerging technologies, and broaden the participation of underrepresented groups in science and engineering. Over 350,000 unique visitors have accessed Cornell Virtual Workshop training on programming languages, parallel computing, code improvement, and data analysis. The platform supports learning communities around the world, with code examples from national systems such as Frontera, Stampede2, and Jetstream2.
The documentation provides an overview of using Pegasus, a workflow management system, on ACCESS resources for high throughput computing (HTC) workloads, covering logging in, workflow creation, resource configuration, and monitoring options.
This workshop series introduces the essential concepts in deep learning and walks through the common steps in a deep learning workflow from data loading and preprocessing to training and model evaluation. Throughout the sessions, students participate in writing and executing simple deep learning programs using Pytorch – a popular Python library for developing, training, and deploying deep learning models.
DARWIN (Delaware Advanced Research Workforce and Innovation Network) is a big data and high performance computing system designed to catalyze Delaware research and education
Learn how to use Linux commands in a python script. Specifically, learn how to use the subprocess and os modules in python to run shell commands (which run Linux commands) in a python script that is run on a cluster.
This Udacity article listed the most frequently used R packages for data science and statistics. For each package, the article provided the link to its official documentation. It will be a great start point if you want to start your data science journey in R.
Open OnDemand is an easy-to-use web portal that lets students, researchers, and industry professionals use supercomputers from anywhere. It is installed on supercomputing resources at hundreds of sites. By eliminating the need for client software or command-line interface, Open OnDemand empowers users of all skill levels and significantly speeds up the time to their first computing.
Matplotlib is a comprehensive library for creating static, animated, and interactive visualizations in Python. It makes analyzing and presenting your data extremely easy and works with Python which many people already know.
Geocoding is the process of taking a street address and converting it into coordinates that can be plotted on a map. This conversion typically requires an API call to a remote server hosted by an organization/institution. The remote server will take the address attributes provided by you and the remote server will compare it to the data it contains and return a best estimate on the coordinates for that location.
There are many geocoding services available with different world coverages, quality of result, and set different rate limits for access. For R, a package called "tidygeocoder" provides an easy way to connect to these different services. As an additional benefit, their documentation provides a good summary of geocoding services available and links to their documentation. The link to the documentation for gecoding services accessible by "tidygeocoder" is provided below.
For Python, geopy package is a library that provides connection to various geocoding services. The link to the documentation for this package is also included below.
This workshop focuses on developing an understanding of the fundamentals of attention and the transformer architecture so that you can understand how LLMs work and use them in your own projects.
This workshop will go into the different ways python packages can be managed in a cluster environment using conda and python virtual environments both in batch mode from the command line and with Jupyter Notebooks and Jupyter Lab on the cluster. The examples will be run on the GMU HOPPER Cluster.
Machine learning is becoming increasingly important in field with large data such as astrophysics. AstroML is a Python module for machine learning and data mining built on numpy, scipy, scikit-learn, matplotlib, and astropy allowing for a range of statistical and machine learning routines to analyze astronomical data in Python. In particular, it has loaders for many open astronomical datasets with examples on how to visualize such complicated and large datasets.
Warewulf is an operating system provisioning platform for Linux that is designed to produce secure, scalable, turnkey cluster deployments that maintain flexibility and simplicity. It can be used to setup a stateless provisioning in HPC environment.
Purdue University is the home of Anvil, a powerful supercomputer that provides advanced computing capabilities to support a wide range of computational and data-intensive research spanning from traditional high-performance computing to modern artificial intelligence applications.
Self-paced tutorials on high-end computing topics such as parallel computing, multi-core performance, and performance tools. Other related topics include 'Cybersecurity for End Users' and 'Developing Webinar Training.' Some of the tutorials also offer digital badges. Many of these tutorials were previously offered on CI-Tutor. A list of open access training courses are provided below.
Parallel Computing on High-Performance Systems
Profiling Python Applications
Using an HPC Cluster for Scientific Applications
Debugging Serial and Parallel Codes
Introduction to MPI
Introduction to OpenMP
Introduction to Visualization
Introduction to Performance Tools
Multilevel Parallel Programming
Introduction to Multi-core Performance
Using the Lustre File System
A guide for Duke OIT on how to advise users on using ACCESS and allocation credits to jetstream 2 for Duke University members. This can be used for non Duke members. Assumes the reader has basic knowledge of ACCESS.
Snakemake is a powerful and versatile workflow management system that simplifies the creation, execution, and management of data analysis pipelines. It uses a user-friendly, Python-based language to define workflows, making it particularly valuable for automating and reproducibly managing complex computational tasks in research and data analysis.
Numpy is a python package that leverages types and compiled C code to make many math operations in Python efficient. It is especially useful for matrix manipulation and operations.
The research paper provides an overview of various datasets that have been used to study fairness in machine learning. It discusses the characteristics of these datasets, such as their size, diversity, and the fairness-related challenges they address. The paper also examines the different domains and applications covered by these datasets.
CAC summer student employee Jeff Lantz describes his experiences in running the WRF weather forecasting application in the public cloud. He compares the major cloud providers and some container-based deployment technologies that are available on each, with a particular emphasis on Docker and Kubernetes. Since WRF is a computationally intensive numerical simulation, Jeff had to pay special attention to certain HPC characteristics of the code, such as the need to launch multiple communicating MPI processes on one or more cloud instances, and the need to set up an NFS file server to satisfy I/O requirements.