Faculty Development Program in Super Computing

The Faculty Development Program in Super Computing is designed to equip educators with essential knowledge and practical skills in high-performance computing (HPC). Through expert-led sessions and hands-on demonstrations, participants will explore supercomputing architectures, parallel programming models, and optimization techniques that enable large-scale scientific and engineering computations. This program empowers faculty to confidently integrate HPC concepts into teaching, research, and academic projects.

By the end of this program, you will be able to:

• Understand the fundamentals of supercomputing and HPC ecosystems.
• Explain cluster architectures, processors, memory hierarchies, and interconnects.
• Use Linux-based environments for HPC workflows and system navigation.
• Apply parallel programming concepts using MPI, OpenMP, or CUDA.
• Optimize code performance for large-scale computation.
• Execute computational tasks on HPC clusters and analyze performance results.
• Integrate supercomputing modules into academic curricula and research activities.

• What is supercomputing?
• Applications in science, engineering, and AI
• Overview of HPC systems in India and worldwide

• Nodes, processors, cores, and accelerators
• Memory hierarchy and storage systems
• High-speed interconnects and scheduling systems

• Command-line essentials
• File systems and resource management
• Shell scripting for automation

• Concepts of parallelism: task, data, and pipeline
• Parallel programming models
• MPI basics: message passing, processes, communication
• OpenMP basics: shared memory, parallel regions, synchronization

• Introduction to CUDA
• Writing basic GPU kernels
• CPU vs GPU performance considerations

• Profiling tools and techniques
• Load balancing and scalability analysis
• Improving computation efficiency

• Running jobs on an HPC cluster
• Writing and executing parallel programs
• Evaluating speedup and computation efficiency

• Operating System: Linux (Ubuntu/CentOS)
• Programming: Python, C/C++ (for HPC modules)
• Frameworks: MPI, OpenMP, CUDA (optional)
• Tools: Slurm scheduler, profiling tools (gprof, perf), shell scripting

• Faculty members teaching computer science, engineering, AI, physics, or related domains
• Researchers requiring high-performance computation for projects
• Educators looking to integrate HPC concepts into curriculum
• Professionals seeking to understand parallel computing workflows

Upon successful completion, participants will receive a Certificate in Super Computing, demonstrating their competency in HPC concepts and practical parallel programming skills.