The COMPLECS CyberTraining program addresses the gap between traditional developer-centric high-performance computing training and the practical needs of modern researchers. Using a three-tier curriculum, the program has impacted more than 1,700 participants in its first 30 months, hosting 43 webinars and 5 immersive, multi-day workshops. Here, we review the program's growth and offer insights for cyberinfrastructure training and improved strategies for user engagement and resource dissemination.
The CI-PIVOT project, a collaboration between Indiana University, Purdue, and Mississippi State, addresses the gap between complex HPC/AI resources and domain researchers by transforming science experts into embedded Cyberinfrastructure Professionals (CIPs). Through an 18-month cohort model, this project establishes a scalable framework for professionalizing the CIP role, providing insights into training that can be generalized across the broader research ecosystem.
This paper addresses the preliminary findings of the discovery survey conducted to establish an initial picture of High-Performance Computing (HPC) resources across Africa. The pilot survey yielded 51 complete submissions from 23 countries, revealing a landscape far more diverse - and in many cases more capable - than previously documented. It also revealed several previously unknown systems, underscoring the value of coordinated continental data-gathering. Simultaneously, the survey exposed methodological and engagement-related limitations that must be addressed before a reliable, long-term HPC catalogue can be effectively realised. We reflect on the execution of the preliminary survey, examining issues such as inconsistent data quality, uneven community participation, gaps in technical detail, and the challenges of surfacing HPC assets in highly resource-constrained environments. Drawing on these observations, we outline the key lessons learned and define the plans for a more robust and targeted follow-up survey. Our aim is to refine the methodology, strengthen community coordination, and build toward a validated, sustainable framework for representing HPC capacity across Africa. Throughout the manuscript, we acknowledge the survey's limitations and use the preliminary dataset to present a snapshot of the current scientific computing infrastructure landscape in Africa.
This paper presents a dual-sided software ecosystem for the NSFfunded ACCESS project, built to streamline the onboarding of heterogeneous HPC resources. Using user-centered design and cognitive frameworks, the application replaces fragmented, high-load processes with a unified workflow that links Resource Providerfacing onboarding pathways to staff verification. Its UI/UX, built around a roadmap-badge-task model, provides a scalable approach for coordinating multi-actor HPC platforms and governance-intensive environments.
Advanced system administration tools, particularly tools that are specifically designed for, and have direct knowledge of the local HPC environment, can greatly improve the productivity of SysAdmins, but often require in-house development efforts beyond the capabilities of many HPC centers. However, recent advances in AI-coding are making it possible for HPC centers with limited development resources to create custom SysAdmin tools quickly and easily. This paper describes our use of AI-coding (via Claude Code) to develop two custom SysAdmin tools (a Cluster Explorer and a Realtime Chatbot) designed to meet our specific needs and tailored to our local HPC environment. We describe our experience with AI code generation, the types of prompts we used to successfully create two highly useful tools in a very short amount of time, and lessons learned.
