This is a recording of a 90-minute virtual workshop on how to use NCSA's Delta supercomputer. It covered Delta's advanced architecture designed to support a wide range of scientific domains, particularly those that benefit from GPU acceleration such as machine learning. Hands-on exercises taught how to access Delta, manage the user environment, and run jobs and software. The content is suitable for researchers and students new to working with a cluster, and who want to know more about Delta.

This is a recorded webinar about programming NVIDIA GPUs using standard programming languages such as C++, Fortran, and Python. It covered how writing parallel-first code enables applications to run on multicore CPUs or GPUs with a single source code. It also described how to mix compiler directives and CUDA with standard languages to get productive, portable, and performant code.

This self-paced tutorial guides you through using Delta for running scientific application code. You'll learn how to access Delta, manage and edit files, set up your software environment, and transfer files. The tutorial includes a hands-on exercise where you'll run an example application. Finally, you'll explore how to use Open OnDemand on Delta. Even if you're not a Delta user, this tutorial offers valuable insights into using an HPC cluster for scientific computing.

A short video explaining the difference between interactive jobs and partitions and how to use them.

This is a recording of a training webinar given by Scot Breitenfeld (HDF), Gerd Heber (HDF), and Hariharan Devarajan (LLNL) on 3/12/2024. It included two presentations about optimizing, profiling, and benchmarking input/output operations using HDF5 and DLIO.

NCSA is hosted a hands-on AI training series comprised of ten sessions (beginning October 26, 2023) to enable researchers to become proficient in using AI techniques on modern supercomputers. This series was taught by experienced AI practitioners and covered basic, intermediate, and advanced AI topics. This page contains the archived materials from these sessions.