A three-and-a-half-day online training titled Accelerating Your Applications with AMD GPUs will be hosted by AMD and NCC Poland from 2 to 4 December 2025, as part of the EuroCC project.
This Zoom-based training, featuring hands-on exercises in each module, is designed for both academia and industry. It targets scientists and researchers with HPC experience, application developers, and HPC practitioners seeking to port, optimize, and profile their codes on AMD systems.
Day 1 is dedicated to System Administration, ML/AI on AMD GPUs, AI Workflows, and AI Programming assistants, covering foundational concepts and practical applications. Day 2 topics will include Python for GPU, an introduction to HIP and ROCm, OpenMP offloading, and GPU-aware MPI, providing essential skills for GPU computing. Day 3 will focus on Optimization and profiling techniques, PyTorch optimization with Large Language Models, Neural Operators, AI Surrogates, and detailed PyTorch profiling, offering comprehensive insights into performance tuning and advanced AI modeling.
Participants should possess some knowledge of GPU and/or HPC programming, along with a general understanding of computer hardware, operating systems, and at least one HPC programming language such as C, C++, Fortran, or Python.
