- This event has passed.
AI WebCafe: How can HPC technologies help AI?
March 17 @ 15:00 - 16:00
The AI Web Café provides the users with possibilities to participate virtually in live sessions with experts on specific AI domain or areas of interests. FocusCoE project will participate with the following three CoEs: RAISE, CoeC, and HiDALGO.
This session aims to bring together the Centers of Excellence in AI sectors to explain how HPC technologies can help AI. Several case studies will be presented on how AI is leveraging HPC technologies to solve real-live problems. Opportunities and challenges in the AI sector will be discussed, how the Centres of Excellence can help to tackle them thanks to HPC technologies and which opportunities are emerging. It is a good opportunity to get a common understanding of the ongoing activities within the different CoEs and foster collaboration opportunities, including education and training offered by Centres of Excellence.
The following session aims to bring together the Centers of Excellence in AI sectors to discuss about the opportunities and challenges in the sector, how the Centres of Excellence can help to tackle them thanks to HPC technologies and which opportunities are emerging. It is a good opportunity to get a common understanding of the ongoing activities within the different CoEs and foster collaboration opportunities. Register now!
1. FocusCoE Introduction
- CoE RAISE – AI in Scientific Workflows on High-Performance Computers (Andreas Linterman) – 15 min
The presentation will provide an overview of the challenges tackled by CoE RAISE and bring examples from two use cases with simulation and data-processing background.
- CoEC – AI for future energy system (Temistocle Grenga) – 15 min
The presentation will provide an overview of CoEC and to show how AI is allowing to overcome problems from scientific applications that were not possible to attack with a theoretical models
- HiDALGO – Generating directed Social Network Graphs (Christoph Schweimer) – 15 min.
The Social Networks (SN) Use Case in the HiDALGO project simulates the spread of information in SN such as Facebook or Twitter. Since the user data in these networks is highly sensitive and must be carefully taken care of, we developed a method to synthesize social network graphs, replicating several topological and algorithmic graph features on which information dissemination can be simulated.
3. Panel Discussion
Temistocle Grenga (CoEC)
Temistocle Grenga is a group leader and researcher at the Institute for Combustion Technology (ITV) at RWTH Aachen University (Germany) since 2018. Previously he was at the University of Notre Dame (USA), where he received his PhD degree in aerospace and mechanical engineering, and at Princeton University (USA). His work focuses on numerical simulation and modeling of multiphase turbulent reacting flows. He coordinates ITV activities at CoEC.
Andreas Linterman (RAISE)
Lintermann coordinates CoE RAISE since 2020 and leads the Simulation & Data Lab “Highly Scalable Fluids & Solids Engineering” at the Jülich Supercomputing Centre, Forschungszentrum Jülich, Germany, since 2014. He received his diploma degree in computer science and his PhD degree in engineering from RWTH Aachen University, Germany. His group develops multi-physics codes in the field of engineering and combines them with novel AI technologies making use of modular supercomputing architectures.
Christoph Schweimer (HiDALGO)
Christoph Schweimer joined the Know-Center as a Data Scientist in 2019 to work in the HiDALGO project on AI-related tasks which support the workflow of the use cases. He received his M.Sc. in mathematics from the University of Salzburg in 2017.
Register now here: https://www.ai4media.eu/aicafe/ai-cafe-how-can-hpc-technologies-help-ai/