Deep learning powers a broad range of applications, from autonomous cars and image generation to voice assistants. The successful training of deep learning models necessitates the use of graphics processing units (GPUs). To facilitate a quick and easy start with our GPUs, we’ve compiled a selection of materials designed to expedite your learning process. Simply follow the instructions below to explore the Nvidia A100 generation GPU on our cluster Grete!

Steps

1. Get access: Create an NHR account at https://zulassung.hlrn.de. This is free for all researchers based in Germany. Is this your first time on any cluster? Check out our bonus material on cluster concepts.

2. Acquire material: Get the freely-available course content at https://gitlab-ce.gwdg.de/dmuelle3/deep-learning-with-gpu-cores . The theoretical part is contained in the slides while the practical part is under code.

3. Practical session: Follow along to bring the workflow of the Gitlab repository to the cluster by watching the video series at https://www.youtube.com/playlist?list=PLvcoSsXFNRblM4AG5PZwY1AfYEW3EbD9O .

What’s included in the course?

• A brief introduction to the core principles of deep learning and how GPUs work during the theoretical session.
• Hands-on learning with our interactive tutorial that guides you to bring a deep-learning workflow to the cluster.
• Examples of how to use the cluster scheduler efficiently.
• A brief introduction to monitoring GPU usage.

Author

Dorothea Sommer