Shaping the future workforce: Challenges and lessons learned in HPC education from national labs and computing centers

UNCLASSIFIED Shaping the future workforce: Challenges and lessons learned in HPC education from national labs and computing centers Patrick Diehl, Ying Wai Li, Christoph Junghans, John K. Holmen, Elijah MacCarthy, Suzanne Parete-Koon, Yun (Helen) He, Rebecca Hartman-Baker, Charles Lively, Kevin Gott, Lipi Gupta, Kristina Streu, Yasaman Ghadar, Paige Kinsley, Jane Herriman, Erik W. Draeger, Victor Eijkhout, and Susan Mehringer LA-UR-25-29400 (https://doi.org/10.5281/zenodo.17603228) Managed by Triad National Security, LLC, for the U.S. Department of Energy’s NNSA. UNCLASSIFIED November 13, 2025 UNCLASSIFIED Motivation 🚀HPC is everywhere! 🚀 •Domain scientists use HPC resources to run simulations •Machine learning needs HPC resources for training •Data science requires HPC resources to process big data sets •Many more examples 🚧HPC is taught “not everywhere”! 🚧 •We need to educate machine learning scientists, data scientists, and domain scientists in HPC •We need to teach proficient HPC users new technologies, e.g. new hardware and software Our group is committed to shaping the future HPC workforce—together, through collaboration. UNCLASSIFIED November 13, 2025 | 2 UNCLASSIFIED Outline Related Efforts This Effort Multi-Center Collaborations Challenges and Lessons Learned: Overview Conclusion and Outlook UNCLASSIFIED November 13, 2025 | 3 UNCLASSIFIED Related Efforts Other active working groups •ACM SIGHPC Education Chapter •CaRCC AI Facilitation Materials Working Group •ADAC Training, Outreach, & Workforce Development Working Group •US-RSE Education & Training Working Group Collection of materials •ACCESS-CI Events and Training •HPC-ED, a metadata catalog of events and materials Exascale Computing Project (ECP) •Sustainable Research Pathways for HPC (SRP‑HPC) •HPC Workforce Development & Retention Action Group (HPC‑WDR) Links to all of them are provided in the paper! UNCLASSIFIED November 13, 2025 | 4 UNCLASSIFIED HPC Training Working Group •During SC 2023, staff from three national laboratories met to compare training efforts •As a result of these discussions, a monthly meeting was established for HPC centers to share their experiences, practices, ideas, and opportunities for collaboration •The group has continued to meet monthly and in 2025 formalized themselves as the HPC Training Working Group 📨If you are interested to join the working group, please contact John Holmen [email protected] UNCLASSIFIED November 13, 2025 | 5 UNCLASSIFIED Participants •Los Alamos National Lab, •Oak Ridge National Lab, •NERSC at Lawrence Berkeley National Laboratory, •Lawrence Livermore National Laboratory, •Argonne National Laboratory, •Cornell Center for Advanced Computing, •Texas Advanced Computing Center, •Purdue University. ⏱️Monthly meetings on the 4th Tuesday of the month at 1:00 PM MT / 3:00 PM ET / 2:00 PM CT / 12:00 PM PT are open to all with new centers welcome to the mix. UNCLASSIFIED November 13, 2025 | 6 UNCLASSIFIED Available Rraining I Training Topics LANL LLNL NERSC ORNL TACC ALCF CAC Parallel Programming 3 4 6 11 2 3 8 Debugging – 2 3 1 1 1 1 Optimization 1 – 2 – – 3 1 Profiling 1 2 3 2 1 2 1 Visualization 1 2 2 1 – 2 4 C++ 1 – – 1 2 – 1 Fortran - – 2 1 – – 1 Python 1 2 3 7 – 2 6 Performance Portability 3 2 4 4 – 2 – GPU programming 1 – 3 3 2 2 1 Storage 1 – 2 2 – 2 – UNCLASSIFIED November 13, 2025 | 7 UNCLASSIFIED Available Rraining II Training Topics LANL LLNL NERSC ORNL TACC ALCF CAC Job schedulers 1 4 – 2 – 3 1 Build systems 1 2 – – 1 2 – Module systems – – 1 – – 2 – Spack 1 2 – – – – – Center-specific Resources 1 8 4 2 3 3 6 Popular OSS - 4 2 2 – 3 – Workflows 1 2 1 2 1 2 – Containers 2 – 2 5 1 1 – Quantum 1 – 3 8 – – – AI, ML, and Data Science 3 2 6 10 – 6 2 UNCLASSIFIED November 13, 2025 | 8 UNCLASSIFIED Multi-Center Collaborations I Cross-Center Advertising •Training collaborations to host the same event across multiple centers. (Example: ParaView training) •Slack channels, Google Spreadsheets, etc. are used to share our events, which can start in the planning stage. •Events hosted by one center and open to other centers, the other centers mirror event pages and advertise in their newsletter. (Example: NVIDIA training) Intro to HPC Bootcamp •The Introduction to High-Performance Computing (HPC) Bootcamp is a one-week immersive program designed to introduce STEM students to fundamental HPC concepts and their application to solving complex scientific and energy-related challenges. UNCLASSIFIED November 13, 2025 | 9 UNCLASSIFIED Conclusion and Outlook Conclusion •Critical role of collaboration between national labs and computing centers in shaping the future of HPC workforce. •Collective effort in context of rapid advancements in artificial intelligence (AI), where HPC serves as a foundational enabler. •AI technologies continue to evolve, the demand for a skilled HPC workforce will grow accordingly, making coordinated initiatives essential for addressing future scientific and technological challenges. Outlook •Sustain and strategically expand the current collaboration and organize a BoF at SC 26 to get national labs and computing centers not yet participating included. •To further enhance the effectiveness of our efforts, we will explore mechanisms for improved organization and sharing of training materials. UNCLASSIFIED November 13, 2025 | 16 UNCLASSIFIED We’d love to hear your thoughts or experiences on this. 📨If you are interested to join the working group, please contact John Holmen [email protected] *Obligatory generative AI image generated by ChatGPT 5 🤖 UNCLASSIFIED November 13, 2025 | 17