Building a Sustainable Intro to HPC Bootcamp

Building a Sustainable Intro to HPC Bootcamp Ayesha Shafiuddin Argonne Leadership Computing Facility Argonne National Laboratory Lemont, USA [email protected] Paige Kinsley Argonne Leadership Computing Facility Argonne National Laboratory Lemont, USA [email protected]v Kristy Henmueller Argonne Leadership Computing Facility Argonne National Laboratory Lemont, USA [email protected] Abstract—High-performance computing (HPC) is critical to advancing STEM research, yet undergraduate curriculum exposure to HPC remains limited, often only offered in graduate-level courses, if offered at all. The Argonne Introduction to HighPerformance Computing Bootcamp addresses this need through a one-week, immersive program, introducing participants to core HPC concepts and computational tools. As part of the bootcamp organizing team, the Argonne professional career intern’s role focused on creating a sustainable framework for the bootcamp’s annual delivery by documenting processes, developing reusable templates, and refining evaluation methods. Additional efforts included onboarding peer mentors, ensuring data quality in evaluations, and capturing both quantitative and qualitative measures of the program’s impact. Through these strategies, the bootcamp aims to expand accessibility, improve participant preparedness for HPC-enabled research, and cultivate a more diverse and skilled HPC community. Index Terms—HPC, workforce development, education, training I. INTRODUCTION With the rise of artificial intelligence and machine learning, the world of high-performance computing (HPC) has been inundated with a wide spectrum of projects investigating various research needs. More than ever before, the Department of Energy’s laboratory ecosystem requires a workforce that can rapidly adapt to the increased HPC needs of many crossdisciplinary STEM fields. However, current undergraduate curricula often lack HPC course material, thereby widening the gap between students and the essential skills necessary to thrive in a rapidly HPC-evolving world. To address these needs, the Department of Energy’s Exascale Computing Project (ECP) Broadening Participation Initiative established three complementary thrusts to recruit and retain the next generation of the HPC workforce [2]. One of these thrusts was the pilot Introduction to HPC Bootcamp, run in August 2023, and organized and hosted in collaboration with Argonne National Laboratory (ANL), Oak Ridge National Laboratory (ORNL), Lawrence Berkeley National Laboratory (LBNL), and Sustainable Horizons Institute (SHI). The goals of the bootcamp focused on bringing together a diverse set of STEM undergraduates and early-career participants for an accessible introduction to HPC [1]. Expanding on the 2023 pilot bootcamp, the Argonne Leadership Computing Facility (ALCF) hosted the 2025 Intro to HPC Bootcamp 1, from August 111https://intro-hpc-bootcamp.alcf.anl.gov/ Fig. 1. Iterative process for bootcamp evaluation and improvement. 15, 2025, organized in partnership with Oak Ridge Leadership Computing Facility (OLCF), and the National Energy Research Scientific Computing Center (NERSC). While the bootcamp is open for everyone to apply to, only 60 participants were selected. These participants were divided into 12 groups of 5 so that the ratio of student to teacher remains low to ensure that each participant receives sufficient individual attention. One of the challenges of meeting the needs of a rapidly evolving HPC world is continuing to train and educate the workforce on the most relevant topics in HPC and AI. While the 2023 Intro to HPC Bootcamp successfully addressed these needs, there were opportunities to further improve the experience. By evaluating and analyzing the 2023 survey results, we identified ways to refine the planning process for future bootcamps. The primary focus of the Professional Career Internship (PCI) intern was to address these improvement opportunities, building a more sustainable bootcamp by documenting essential workflows, improving evaluation tools, and refining an iterative growth-focused bootcamp planning process. In addition, the PCI internship experience offered both technical and professional development opportunities through exposure to HPC concepts and a more expansive peer network. As a result of improvements in bootcamp planning workflows, the 2025 Intro to HPC Bootcamp was a smoother experience for both organizers and participants. Unless otherwise stated, the rest of this paper describes the activities directly guided by or supported by the PCI intern. Fig. 2. Relational hierarchy for different roles in the bootcamp. II. METHODS The approach to optimizing the planning process of the 2025 Intro to HPC Bootcamp included reviewing the 2023 Intro to HPC Bootcamp evaluation data, innovating upon the current bootcamp planning processes, documenting the new processes, and collecting data in the current iteration of the bootcamp (Figure 1). The evaluation data from the 2023 bootcamp was curated in the form of post-bootcamp survey that included both qualitative and quantitative data collected from participants, peer mentors, and project leads. As the structure of the bootcamp is oriented around the HPC-related projects and tools to teach HPC, much of the survey data reflected how those sessions could be improved. Additionally, feedback provided insight into participant and peer mentor onboarding processes. To document the new or improved planning processes of the bootcamp, we created a master document that functioned like a master key. The design is built with future bootcamp planning in mind, storing logistic email templates, documents that contain essential workflows, and various other content pertinent to the bootcamp’s success. Furthermore, we updated both a pre-bootcamp survey as well as a post-bootcamp survey, based on 2023 evaluation tools, that measured growth regarding HPC skills, programming skills, and a variety of open-ended questions to collect qualitative feedback. III. RESULTS AND DISCUSSION The structure of the bootcamp is as follows: 60 participants with 5 participants per group, 12 peer mentors to support project groups, and 7 projects with at least one project lead per project (Figure 2). The 2023 evaluation data revealed that participants and peer mentors both needed much more support leading up to the bootcamp. Specifically, participants asked for more technically related resources to ensure their coding skills would not limit their participation. This was supported by peer mentor feedback that noted participants with weak coding skills disengaged and struggled to keep up during the bootcamp. Peer mentors requested more time with the project materials before the bootcamp, along with specific guidance on best practices to support participants. An essential approach that we worked on during the 2025 iteration was the development of the peer mentor program. Based on feedback from 2023, we onboarded the peer mentors two months before the bootcamp, introducing them to their respective project leads and clarifying expectations about their role during the bootcamp. To actively address challenges or surprises during the bootcamp, daily debriefs were set up with peer mentors, providing a constant flow of information about the participants’ learning experiences that could be actively implemented throughout the week. This allowed presenters to actively adjust their talks based on feedback from previous days. Though we have not yet parsed through the feedback of the post-bootcamp surveys, we are planning to see what themes arise in participant feedback, and whether they match the pattern of feedback received from peer mentors during the daily debriefs. Another unique opportunity that presented itself was an interest from some peer mentors in the role of the PCI intern. This presents a unique opportunity to build the pipeline of developing a program from participant to peer mentor to a PCI intern who supports the bootcamp at a higher level of organization. This forges a path and introduces a natural progression for undergraduate students or early career professionals to enter their career in the world of HPC, giving them a holistic view of training, enabling, and sharing HPC. IV. CONCLUSION AND NEXT STEPS Building a sustainable bootcamp allows for maximum impact and a smooth learning experience for participants. To build a sustainable bootcamp means 1) reviewing previous survey evaluation data, 2) innovating upon the current planning process of the bootcamp, 3) documenting the planning process, and 4) collecting new feedback on the current iteration of the bootcamp. This evaluation cycle is critical in building the most effective bootcamp to meet the needs of priming the next generation of HPC professionals across a spectrum of STEM disciplines. In addition, including a PCI intern in the bootcamp planning process provides bandwidth to organizers to continually help improve the iterative evaluation cycle, as well as build a pipeline to further develop early-career professionals in the world of HPC. The world of HPC is not solely limited to researchers and scientists who use HPC systems. Rather, the entire ecosystem of HPC relies on the expertise of those who build systems, enable research, and share the skills of HPC to a workforce in need. The Intro to HPC Bootcamp is one facet of this world that helps share HPC to a workforce that is growing rapidly. Improving the bootcamp iteratively will only lead to a more meaningful and impactful bootcamp for future HPC professionals. ACKNOWLEDGMENT I am eternally grateful to my mentors for all their guidance and support. This work was supported by the Argonne Leadership Computing Facility, a DOE user facility under the U.S. DOE Office of Science-Advanced Scientific Computing Research Program, under Contract No. DE-AC02-06CH11357. REFERENCES [1] LEUNG, M. A., CAHILL, K., HARTMAN-BAKER, R., KINSLEY, P., CURFMAN MCINNES, L., PARETE-KOON, S., RAMPRAKASH, S., ABRAHAM, S., BEACH BARRIER, L., CHEN, G., ET AL. Intro to hpc bootcamp: Engaging new communities through energy justice projects. Journal of Computational Science Education 15, 1 (02 2024). [2] PARETE-KOON, S., LEUNG, M. A., RAMPRAKASH, S., AND MCINNES, L. C. Exascale computing project’s broadening participation initiative. The Journal of Computational Science Education 14 (July 2023), 53–54.