Speaker
Description
Introduction to PHYSnet and UHH GPU-Ressources
The PHYSnet-RZ as well as the UHH-RRZ each provide compute resources, including GPUs, for you. This talk will provide you with an overview of the available resources, their differences and individual target use-cases and how-to access them.
We will focus on GPUs in particular, as this type of accelerator hardware currently receives a lot of attention due to its particular benefits for increasingly popular machine-learning use-cases. We will discuss GPU's potential in speeding up your computations along with which use-cases do or don't benefit from including GPUs and why that's the case.
The talk will include a short live demonstration on how-to access the PHYSnet compute cluster, setup a common base-environment for machine learning tasks with Python and run a tiny example compute job using GPUs.
Chair: Dieter Horns
