DYNAVERSE - Our Dynamic Universe
Dynaverse is a new Cluster of Excellence initiative between the Universities of Cologne and Bonn, the Forschungszentrum Jülich, the Max Planck Institute for Radio Astronomy, the German Aerospace Center and the Heidelberg Institute for Theoretical Studies. It is a world-class hub of expertise in radio astronomy, lab-experiments, simulations, and machine learning / artificial intelligence. Dynaverse brings together researchers from Astrophysics, Computer Science and Mathematics. State-of-the-art research facilities to which our team has priority access combined with the expertise in high-performance computing, and data-analytics will generate new ground-breaking opportunities.
The dynamic evolution of our Universe is regulated by nonlinear physical processes that occur on vastly different timescales, from fractions of a second to billions of years. How are they interconnected to shape the Universe? The proposed Cluster of Excellence “Our Dynamic Universe” (Dynaverse) will deliver a breakthrough in our physical understanding by addressing three core Astrophysical Research Challenges (ARCs):
- ARC1: Time-Lapse Astronomy
- ARC2: Slow-Motion Astronomy
- ARC3: Cosmic Twists
- The SUE: The Shared Universe Engine
The Bonn-Cologne area is a world-class hub for astrophysics. It builds on a holistic approach with four complementary Pillars of expertise. P1: developing state-of-the-art detectors and instruments for international telescopes; P2: leading numerous large-scale observational programmes based on open and guaranteed observing time; P3: operating top-tier laboratories for astrophysics; and P4: simulating the evolution of planets, stars, and galaxies on high-performance computers. State-of-the-art facilities, to many of which our team has prime access as a major partner, offer groundbreaking opportunities. These include upcoming radio telescopes delivering up to zettabytes of data per year, complementary multi-wavelength facilities, and the first European Exascale computer JUPITER at Forschungszentrum Jülich.
Optimally exploiting the novel opportunities in the transformative era of Big-Data-Astrophysics promises breakthroughs. To succeed, overcoming the extreme challenges of the unprecedented big and heterogeneous data is crucial. In this new era, the astrophysical community needs innovative methods for discoveries. Dynaverse addresses this by combining experts in Astrophysics, Computer Science, and Mathematics to establish a fifth Pillar P5: Astroinformatics. Through P5 we will leverage AI and machine learning to develop new methods for astrophysics. All our data and insights will be publicly accessible in the Shared Universe Engine (SUE), a smart, open-source workflow and processing hub. The SUE will be a machinery of discovery and a collaborative workspace for the Dynaverse team and the community.
The Jülich Supercomputing Centre is honored to be part of the Dynaverse Excellence Cluster initiative and contributes with its expertise in cutting-edge AI technology development, algorithm and AI training scalability, high-performance computing, and big data analytics. The PIs Prof. Susanne Pfalzner (SDL Astronomy and Astrophysics), Dr. Andreas Lintermann (SDL Fluids & Solids Engineering), and the cooperation partner Dr. Stefan Kesselheim (SDL Applied Machine Learning) play key roles in the establishment of Pillar P5: Astroinformatics, contribute to various developments concerning Computational Science questions in the different Pillars, and in the development of SUE.
More information can be found on the official Dynaverse website: https://dynaverse.astro.uni-koeln.de/
Contacts at JSC:
Prof. Dr. Susanne Pfalzner
Head of SDL Astronomy and Astrophysics PI in Helmholtz Information Program 1, Topic 1
- Institute for Advanced Simulation (IAS)
- Jülich Supercomputing Centre (JSC)
Room 319
Dr. Andreas Lintermann
Leader Simulation and Data Lab "Fluids & Solids Engineering" PI in Helmholtz Information Program 1, Topic 1
- Institute for Advanced Simulation (IAS)
- Jülich Supercomputing Centre (JSC)
Room 4018
Prof. Dr. Stefan Kesselheim
Head of SDL Applied Machine Learning & AI Consultant team PI in Helmholtz Information Program 1, Topic 1
- Institute for Advanced Simulation (IAS)
- Jülich Supercomputing Centre (JSC)
Room 3023