Nenad Mijić

Nenad Mijić received his M.S. degree in Physics from the University of Zagreb in 2017. He is currently working on his PhD under the supervision of Davor Davidović, Ruđer BoškovićInstitute, Zagreb. His research interests include code portability and eigenvalue solvers on heterogeneous distributed systems and high performance computing. Often wanders into ML territory as he develops an interest and passion for it. He currently works at the University Computing Centre SRCE, as part of the Advanced Computing Department team.

High Performance Computing:  Empowering the AI

AI algorithms, particularly deep learning models, require immense computational power to train on vast datasets. HPC systems provide the parallel processing capabilities necessary to handle these intensive computations efficiently. Without HPC, the development of complex AI models, especially large language models (LLMs), would be impossible.

Utilizing HPC for AI offers numerous advantages, including reduced training times, the ability to process larger datasets. HPC systems accelerate AI research and deployment, enabling faster iterations and more sophisticated models that can drive business innovation and competitiveness.

University Computing Centre SRCE offers several services for industry users, ranging from HPC systems to dedicated support and expertise for AI research and development. Also, SRCE offers support in accessing a network of HPC systems across Europe to foster innovation and research.

In this workshop, we will present benchmark results to demonstrate the tangible benefits of using our HPC system for AI workloads. Furthermore, we will detail the available HPC systems at SRCE and in Europe and how to access them.