CI Pathways - Spring 2025
Topic outline
-
Are you ready to use cutting-edge cyberinfrastructure (CI) tools to elevate your scientific work? If you are, join CI Pathways (NSF award 2417789) and embark on a transformative journey to empower researchers like you!
Spring 2025 Session Schedule:
The link to each event's Zoom session is available from the Upcoming Events listing (right side of screen).Set your Timezone—The event times shown in the Upcoming events calendar are based on the timezone you have set in your User Profile. To edit your timezone, click on the down arrow next to your user icon in the upper right of the screen. Then select Preferences, Edit Profile. Timezone is one of the fields shown.
-
The CI Awareness pathway covers key elements of how Cyberinfrastructure (CI) is used in scientific research. It consists of four sessions: an overview of CI's essential components and practical applications; the basics of using Jupyter Notebooks,; Machine Learning/Artificial Intelligence and Data Science; and Parallel Computing. There are no expected prerequisites for this pathway other than an interest in learning how cyberinfrastructure can be used in research workflows.
-
The Data Science Pathway consists of four training sessions. The first one teaches how to use the AI-tailored NCSA Delta HPC system which will be used for the hands-on activities. The following three sessions will cover the basics of using HPC tools to manage data from simple databases (such as Pandas, SQL) to massive distributed datasets (using Apache Spark).
Participants in this pathway should be familiar with basic python programming, such as NumPy/SciPy and basic Linux shell commands.
-
This pathway consists of four training sessions covering: how to use the AI-tailored NCSA Delta HPC cluster which will be used for the hands-on activities; the basics of PyTorch (a popular programming framework); signal data processing using convolutional neural networks; and deep learning training using multiple graphical processing units (GPUs). Participants in this pathway are expected to have familiarity with basic Python programming, such as using NumPyand basic linear algebra (matrix-matrix multiplication).
-
The Parallel Computing Pathway consists of four training sessions. The first one teaches how to use the AI-tailored NCSA Delta HPC system which will be used for the hands-on activities. The following three sessions will cover essential parallel programming frameworks and architectures in shared memory (OpenMP) and distributed memory systems (MPI). These are critical tools to scale up scientific research applications and leverage the power of supercomputers at scale using CPUs and GPUs.
-
-
-
-
-