Abstract:

Traditional approaches to learning High-Performance Computing (HPC) and software development languages often remain highly theory-oriented due to limited access to suitable supercomputing infrastructure and interactive learning environments. As a result, learners frequently lack opportunities to gain meaningful hands-on experience with real computational systems. To address this challenge, C-DAC present HPC Tutor, an interactive learning platform designed to bridge the gap between theoretical understanding and experiential practice in both HPC education and computational skill development.

HPC Tutor enables end users from any background to study HPC concepts as well as computational languages while developing practical programming and system-level skills through interactive exercises and guided experimentation in real time environment. The platform helps users develop computational skills through a range of learning modules, including multimedia lessons, quizzes, and guided knowledge checks. These modules introduce concepts such as parallel computing, system usage, and best programming practices. Unlike traditional e-learning systems, to further strengthen hands-on learning, the platform incorporates real-time laboratory modules where users can test commands, run scripts, and experiment with workloads in a guided setting. A browser-based terminal provides secure shell access to the underlying system, enabling learners to become familiar with command-line workflows commonly used in HPC and software development environments.

To improve learner support, we have integrated an in-house built intelligent chat bot that learners can use to have their questions answered. The chat bot leverages an internal knowledge base to understand commands, provide explanations, and offer contextual support while learners work through tutorials and exercises. This approach provides consistent support without requiring reliance on external internet search engines. Integration with Jupyter Notebooks enables interactive coding, data analysis, visualization, and rapid prototyping directly within the learning environment.

In addition to HPC-focused training, HPC Tutor also supports general computational learning, the platform can also be used for programming courses, data analysis exercises, and computational problem-solving workflows. Students can enroll in courses, complete quizzes and assignments, and monitor their progress through personalized dashboards. Administrators can efficiently manage course content and evaluate learner performance.

Initial deployment involving over 300 learners in the Master Training Program and evaluation across more than 1000 users in multiple training programs demonstrate that HPC Tutor significantly enhances learner engagement, improves conceptual understanding, and promotes hands-on experimentation. The platform is currently deployed across four national supercomputing facilities located in Bangalore, Mumbai, Delhi, and Pune, as well as on an in-house tabletop supercomputer, PARAM Shavak, providing learners with access to real HPC environments across diverse infrastructure setups. By integrating structured learning resources, interactive coding environments, intelligent assistance, and direct interaction with computational systems, HPC Tutor offers a scalable and effective platform for modern computational education and HPC skill development..

Keywords:

Technology, Education, Development, High Performance Computing education, Interactive Learning Platform, Virtual Laboratories, Jupyter Notebooks, AI Chatbot, Technology Enhanced Learning, Computational Education.