Descriptions of phenomena such as Big Data and so-called “algorithmic cultures” (Seyfert & Robert, 2017) indicate a shift in how knowledge is produced, as scientific understanding and social reality are increasingly shaped by data-driven models, and algorithmic decision-making processes. Computational procedures have thus become central to knowledge production, with methods from computational and data science established as growing range of disciplines (Neumann et al., 2022). High Performance Computing (HPC) plays a key role in this development, not merely as technical infrastructure but as socio-technical practice that enables complex calculations, large-scale simulations and data-intensive research (Raj et al., 2020; Barelli & Lodi, 2025; Bacchio, 2025).

As the relevance of HPC expands, so do the demands for HPC-related competencies. These competencies are not limited to technical skills but encompass methodological, organizational, and epistemic dimensions that enable effectively participation in data-driven research practices (Serik et al. 2021; Stevens et al., 2024). However, HPC competencies are rarely embedded in bachelor’s and master’s curricula, resulting in fragmented learning strategies, underutilized infrastructures, and unrealized learning potential (Neumann et al., 2022). Researchers are required to engage in self-directed and ongoing competence development (Schicke, 2014), often relying on tutorials, short-term workshops, or OER’s (Mullen et al., 2016). Given the complexity of HPC, such individualized approaches remain insufficient (Raj et al., 2020).

From an adult education perspective, these challenges point to the necessity of collective and socially embedded learning structures with integrated educational opportunities that are accessible to beginners, scalable for advanced users, and transferable across disciplines. In adult education, HPC education represents an underexplored domain where learning opportunities can be designed beyond formal curricula, shaping new disciplinary and research possibilities (Raj et al., 2020). For that, existing experiences from HPC educational programs in higher education indicate that project-based learning and active experimentation with small self-guided projects, low-threshold support for programming, and collaborative exchange foster self-directed competence acquisition yield better learning achievements (Barelli & Lodi, 2025; Raj et al., 2020) by enhancing agency in relation to the world (Holzkamp, 1995). This shifts the focus toward the question of how learners, experts and other stakeholders can be embedded into social structures that systematically enable the exchange of experiences, problem-solving strategies, and practical knowledge. Here, the concept of community, frequently invoked in the HPC education literature (Raj et al., 2020; Barelli & Lodi, 2025; Bacchio, 2025), opens central perspectives.

The hpc.bw (dtec.bw) project addresses this by establishing an HPC Competence Center to facilitate interdisciplinary access and support competence development across diverse user groups: from non-specialists to HPC experts (Neumann et al., 2022). Competence development in hpc.bw is conceptualized as a shared practice rather than an individual endeavour. Drawing on Communities of Practice theory (Wenger, 1998), learning formats enable continuous exchange among learners, experienced users, and experts, systematically sharing practical knowledge. Within the context of digital transformation and data-driven research, this contribution examines hpc.bw through this lens, presenting learning offers and support structures as a heuristic framework.

In the conference contribution, the concept of community will be examined as a central analytical category for understanding learning and competence development in the context of HPC education. The contribution aims to explore how the term community is mobilized, interpreted, and operationalized within the hpc.bw project and related HPC education discourses, and how it can be meaningfully framed from an adult education perspective, as it is widely used but conceptually ambiguous (Zeuner, 2020). Additionally, special attention is given to learning experiences and competence requirements of researchers from fields not traditionally associated with HPC, exploring how educational design can facilitate new research potentials and self–world relations (Koller, 2012) in a data-driven environment.

The study follows an exploratory–reconstructive mixed-methods design that sequentially combines qualitative and quantitative approaches. The methodological approach aims to reconstruct how learning and problem-solving processes are organized, experienced, and supported within the context of HPC-related continuing education. As the first step, internal project documents (including project descriptions, program outlines, and communication materials) were analyzed (Prior, 2009; Bowen, 2009) to reconstruct goals, implicit assumptions, and structural logics of the hpc.bw project. This document-based analysis focused on the internal project logics shaping HPC education and competence development. In addition, a programmatic analysis (Nolda, 2018) of the educational offerings developed within the hpc.bw was conducted, explicitly examining these formats from the perspective of community building, including aspects such as peer exchange, shared practices, and opportunities for sustained participation. These findings were expanded by a literature review of HPC education from an adult education perspective. The review aimed to identify existing concepts of HPC competencies, learning objectives, and qualification requirements and to situate them with broader adult education theory. Together, the document analysis, programmatic analysis and literature review served to develop a conceptual reference framework for the empirical investigation. To empirically contextualize these conceptual assumptions, two standardized quantitative surveys were conducted among participants of HPC-related educational formats within the project framework of hpc.bw. The first survey (N= 37) focused primarily on capturing users’ needs, prior experiences, and perceived barriers in working with HPC infrastructures. The second survey (N= 35) focused on participants’ problem-solving strategies, previous learning pathways, and approaches to knowledge acquisition in the HPC context. Both surveys aimed to reconstruct how participants work with HPC resources, how learning processed are organized, and which forms of support are perceived as relevant. The surveys were conducted as quantitative full surveys with small sample sizes. Collected data included self-assessments of HPC-related skills, experiences with HPC usage, problem-solving strategies, learning media, perceived support, and continuing education needs. Due to the exploratory character and limited sample size, the quantitative analyses primarily served to identify patterns, tendencies, and relational structures rather than statistically generalizable results. The findings informed both the interpretation of learning and working practices in HPC contexts and the further development of competence-oriented and community-based educational offerings within the project. The results were processed and disseminated in the form of scientific posters, contributions to conferences and smaller papers. A journal article will be published by spring 2026.

The findings indicate that central challenges in HPC education lie less in purely technical limitations than in high entry barriers to HPC-related research practices, fragmented learning pathways, and insufficiently embedded support structures. The study reveal that learning and problem-solving in HPC contexts are strongly shaped by how learning opportunities are socially organized and connected to everyday research practices. The results indicate that community-based learning structures, characterized by peer-to-peer exchange, shared methodological approaches and practice-oriented support, are particularly effective in addressing these challenges. Rather than relying on individualized or purely material learning resources, the examined educational formats foster sustained engagement with HPC by embedding competence development within collective practices. In this sense, HPC-related Communities of Practice function as enabling structures that lower entry barriers, support learning processes, and connect technical competencies to concrete research horizons. Furthermore, the findings demonstrate that researchers from disciplines not traditionally associated with HPC can develop new research potentials through the acquisition of HPC competencies, provided that appropriate learning and support structures are available. Competence development emerges not as a linear acquisition of competencies but as a socially situated process shaped by interaction and collective problem-solving strategies; through the theoretical heuristic of Communities of Practice, HPC competencies become visible as shared practices that evolve through participation rather than as isolated individual capabilities. Overall, the study contributes to the theoretical grounding of HPC education as an object of adult education research and empirically illustrates how community-building strategies can support sustainable competence development in data-driven research contexts. It highlights the relevance of Communities of Practice as a conceptual and practica framework for designing educational infrastructure that enable long-term engagement with HPC beyond formal curricular.

Bacchio, S. (2025). High-Performance Computing: Why & How. An introductory guide for getting started with HPC. Computation-based Science and Technology Research Center (CaSToRC), The Cyprus Institute, Nicosia. https://hpc-portal.eu/sites/default/files/2025-08/High-Performance-Computing_-WhyHow.pdf. Barelli, E., & Lodi, M. (2025). The Big Ideas for HPC Education: From Existing Needs in High-Performance Computing Training to Recommendations for Instructional Design. ICSC Observatory. https://doi.org/10.5281/zenodo.17097602. Bowen, G. (2009). Document Analysis as a Qualitative Research Method. In Qualitative Research Journal (2009) 9 (2): 27–40. https://doi.org/10.3316/QRJ0902027. Holzkamp, K. (1995). Lernen: Subjektwissenschaftliche Grundlegung. Campus Verlag. Koller, H.-C. (2012). Bildung anders denken: Einführung in die Theorie transformatorischer Bildungsprozesse. Stuttgart: Kohlhammer. Mullen, J., Arcand, W. R., Bestor, D. A., Bergeron, W. J., Byun, C., Gadepally, V. N., Houle, M., Hubbell, M., Jones, M. S., Klein, A. P., Michaleas, P. W., Milechin, L., Prout, A. J., Rosa, A., Samsi, S. S., Yee, C., Kepner, J., & Reuther, A. I. (2016). Designing a new high performance computing education strategy for professional scientists and engineers. In 2016 IEEE Conference on High Performance Extreme Computing (HPEC) (pp. 1–6). IEEE. https://doi.org/10.1007/s10458-022-09565-7. Neumann, P., Duffek, J. A., Kleinschmidt, J., Leinen, W. G., Breuer, M., Schmidt-Lauff, S., Fink, A., Mayr, M., Firmbach, M., Popp, A., & Auweter, A. (2022). hpc.bw: A supercomputer with competence platform for the universities of the Federal Armed Forces. In D. Schulz, A. Fay, W. Matiaske, & M. Schulz (Eds.), dtec.bw – Beiträge der Helmut-Schmidt-Universität/Universität der Bundeswehr Hamburg: Forschungsaktivitäten im Zentrum für Digitalisierungs- und Technologieforschung der Bundeswehr dtec.bw (Vol. 1, pp. 305–310). https://doi.org/10.24405/14569. Nolda, S. (2018). Programmanalyse in der Erwachsenenbildung/Weiterbildung – Methoden und Forschungen. In R. Tippelt & A. von Hippel (Eds.), Handbuch Erwachsenenbildung/Weiterbildung (6th ed., pp. 433–449). Springer VS. Prior, L. (2009). Using documents in social research. SAGE Publications. Serik, M., Yerlanova, G., Karelkhan, N., & Temirbekov, N. (2021). The Use of The High-Performance Computing in The Learning Process. International Journal of Emerging Technologies in Learning (iJET), 16(17), 240–254. https://doi.org/10.3991/ijet.v16i17.22889. Seyfert, R., & Roberge, J. (Eds.) (2017). Algorithmuskulturen. Über die rechnerische Konstruktion der Wirklichkeit. Transcript. Stevens, C., Anderson, S. M., & Carlson, A. (2024). Integrating high performance computing into higher education and the pedagogy of cluster computing. In Practice and Experience in Advanced Research Computing 2024: Human Powered Computing (PEARC ’24) (Article 106, pp. 1–3). Association for Computing Machinery. https://doi.org/10.1145/3626203.3670588. Zeuner, C. (2020). Community Development and Education. Konzeptionelle Positionierung zwischen Affirmation und Emanzipation. Hessische Blätter für Volksbildung, 22–40. https://doi.org/10.3278/HBV2002W003