HPC Awareness Workshop 2026

About

Under the National Supercomputing Mission (NSM), C-DAC is committed to the continuous upskilling of NSM users. Each year, a new cohort of users is onboarded on NSM systems, resulting in a recurring need to orient and train new users. Many of these newcomers initially face challenges in effectively accessing and utilizing the high-performance computing (HPC) systems deployed under NSM.

To address this requirement, C-DAC is organizing a comprehensive HPC Awareness Workshop aimed at building foundational knowledge in core high-performance computing concepts, tools, and technologies.

The workshop is spread over a period of one month, with two sessions conducted every week on Thursdays and Fridays.

The program introduces essential HPC concepts and provides practical exposure to key technologies such as OpenMP, MPI, GPU programming, and efficient usage of HPC resources using job schedulers. Designed for students, professionals, and researchers, the workshop strikes a balance between theoretical understanding and hands-on practice.

Objectives

• To familiarize new NSM users with HPC systems and environments
• To build foundational knowledge of parallel programming models and tools
• To enhance user confidence in accessing and efficiently utilizing NSM resources
• To enable participants to effectively contribute to research and development goals under NSM

This initiative aims to equip participants with the necessary knowledge and skills to navigate HPC systems efficiently and actively support the objectives of the National Supercomputing Mission.

Topics to be Covered

The following topics will be covered under this training program:

• Introduction to HPC
• HPC environment setup and cluster access
• Linux basics
• SLURM workload manager
• Domains in HPC and HPC application installation using source and SPACK
• Open Multi-Processing (OpenMP)
• Message Passing Interface (MPI)
• Introduction to GPUs and CUDA programming
• OpenACC (Open Accelerators)
• HPC profiling tools
• AI/ML/DL workloads on HPC clusters