Training & Education

This is the first part of a two-part workshop together with AMD to teach the usage of the MI300A GPU which is employed in our VIPER-GPU supercomputer. On four half-days, experts from AMD will present different topics related to this new APU architecture. Together with exercises on the AMD cloud system, you will learn about HiP and the transfer of a CUDA code to HiP, about using OpenMP TARGET directives for GPU porting in a portable way, standard language constructs and the toolsuite that comes with the ROCM system.The second part will follow the week after with a hackathon where you can bring-in your own codes and we will work on them together to reach porting or performance goals on our VIPER-GPU system. Please register for the two events separately. [more]

In this hackathon part of the AMD-GPU workshop, you can bring-in your own codes and port/optimize them on our VIPER-GPU system together with experts from the MPCDF and AMD. If you have already sufficient knowledge of the APU architecture and the GPU paradigm to use for your code, you can participate in this course without the first part.In general, we expect the content of the first part to be known as far as it is relevant for your code, otherwise please also register for the first part of the AMD-GPU workshop separately. [more]

MPCDF in partnership with NVIDIA, is pleased to invite you to a two‑day, on-site only, AI workshop. Over the course of the event, leading NVIDIA specialists and MPCDF experts will deliver high‑impact talks and facilitate networking opportunities—perfect for researchers looking to accelerate scientific discovery with AI. Day 1 dives into the latest concepts for training, fine‑tuning and efficiently deploying Large Language Models, while Day 2 highlights AI‑driven breakthroughs in physics, medical imaging and digital biology. [more]

The Introductory User Tutorial is designed to familiarize you with MPCDF's compute and data services. In the first part, we provide an overview of MPCDF's mission and highlight its key services, along with a step-by-step guide on how to obtain a user account. Moving on, the second talk delves into MPCDF's supercomputers, covering their hardware specifications, software environment, filesystems, and file transfer and archiving workflows. The final part offers a comprehensive introduction to using the Slurm batch system, demonstrating its functionalities through practical examples and job scripts drawn from MPCDF's supercomputers. Throughout the talks, feel free to ask questions via chat. Plus, we offer a dedicated Q&A session afterwards. This tutorial aims to simplify the initial experience for newcomers to MPCDF's services. We conduct it twice a year, typically in spring and autumn, and past presentation slides are available for reference. [more]