HPC for AI Workloads, A Technical Workshop by IIT, Delhi and Altair

May 26, 2023 

First floor, LH 213, Lecture Hall Complex, IIT Delhi, Hauz Khas, New Delhi

High-Performance Computing (HPC) has been a highly effective research and development tool for scientists and researchers in their pursuit of innovation, with a high return on investment compared to other infrastructure tools. As Artificial Intelligence (AI) becomes increasingly integrated into various areas of R&D and innovation, it is crucial for countries to possess the necessary skills, tools, and methods to leverage this technology. While HPC and AI share similar computing environments in terms of infrastructure, their workflows and frameworks are different.

This workshop aims to equip the HPC user community with the skills to harness their existing investment in HPC for AI workloads, avoiding the creation of separate silos for HPC and AI infrastructure. These workloads are often data-intensive, and optimizing application performance is critical, including identifying performance bottlenecks at the storage, network, and I/O levels.

KEY TAKEAWAYS:

• Skill development leadership for the HPC user community.
• Promotion of AI technologies and tools in R&D activities.
• Attraction of top research talent to the institute through showcasing of R&D infrastructure.
• Integrated workflows for HPC and AI.
• Identification and resolution of performance bottlenecks.
• Enhanced proficiency in HPC and AI technologies.
• Increased adoption of AI for improved research outcomes.
• Reputation enhancement as a hub for advanced research and development.
• Optimization of application performance in data-intensive research activities.
• Empowerment of researchers and scientists to leverage HPC for AI workloads.       
  

   

  KEY SPEAKERS

                                                  
• Mr. Sudeep Narayan Banerjee is working at Indian Institute of Technology Delhi as System Architect in Computer Services Centre since Nov 2022 with a total experience of 15+ years in the field of High Performance Computing, Virtualisation, Parallel and Distributed Computing, System and Network Administration, Data Center Management. Mr. Banerjee received his M.Tech in Computer Science and Infrastructure from BITS Pilani and B.Tech in Electronics and Instrumentation Engineering from West Bengal University of Technology, Calcutta. He has designed and commissioned several HPC clusters in former institutions. His research interests include HPC Cluster, GPU/DPU, NFV, Cloud Computing, Security. He loves to teach and is always ready to converse on topics to Physics, Computers, Life and Education.
      

• Mr. Subhasis Bhattacharya has 24 years of experience in mission-critical systems and architecture. He has been instrumental in devising architectures for products associated with High Availability, Disaster Recovery, and High-Performance Computing. He currently heads the PBS, Grid engine, and Accelerator R&D teams worldwide at Altair. His key interests have been in the areas of Scalability and Performance of workload managers for Exascale/HTC, Cloud computing, and Integrating ML workloads with traditional HPC systems.