Scientific symposium on the occasion of the 60th anniversary of RZG/MPCDF

On the occasion of the 60^th anniversary of the foundation of the RZG (now the Max-Planck Computing and Data Facility), the MPCDF hosts a scientific symposium on high-performance computing and data science at its site at the research campus in Garching on October 14th, 2021. Participation is by invitation only.

The event is supported by longstanding industry partners of the MPCDF:

        .

Current Corona Policies

Registered guests are kindly asked to provide proof of "3 G" (short for German geimpft, getestet, genesen - vaccinated, tested, recovered). The 3 G proof shown upon arrival at the lecture hall is also valid for the dinner afterwards.

Program

• 12.30 Admission at TUM lecture hall MI HS 1, Scientific Campus Garchig
• 13.00 Opening: Klaus Blaum (Vice President MPG), Erwin Laure (Director MPCDF)
• 13.30 Sibylle Günter (MPI for Plasma Physics): 60 years computer center Garching - a success story for IPP and MPG
• 13.55 Ignacio Cirac (MPI of Quantum Optics): Quantum Computing and Simulation
• 14.20 Matthias Scheffler (Fritz Haber Institute of the MPG): Paradigm shift by data-centric materials science jointly steered with and enabled by MPCDF
• 14.45 Coffee Break
• 15.10 Claudia Felser (MPI for Chemical Physics of Solids): Topology and Chirality
• 15.35 Moritz Helmstaedter (MPI for Brain Research): Connectomics
• 16.00 Volker Springel (MPI for Astrophysics): The dark and luminous sides of cosmic structure formation
• 16.25 Break
• 16.30 Hans-Joachim Bungartz (Technical University Munich): The new science - it's computational!
• 16.55 Gerhard Hummer (MPI of Biophysics): Computing in times of COVID
• 17.30 Bus shuttle to Munich Olympiaturm
• 18.00 Reception and dinner
• 21.30 Closing

Abstracts

• Prof. Sibylle Günter (Max Planck Institute for Plasma Physics)
  60 years computer center Garching - a success story for IPP and MPG

• Prof. Ignacio Cirac (Max Planck Institute of Quantum Optics)
  Quantum Computing and Simulation
  Quantum computing combines the strengths of Quantum Physics and Information Theory in order to process information in a very efficient way. In fact, it is expected that quantum computers will be able to solve problems that are beyond the capabilities of existing or planned supercomputers. In this talk I will explain how those devices work, review the current efforts to build them, and give some examples of their potential impact. In particular, I will concentrate on simulation problems that appear in the areas of condensed matter physics and chemistry.

• Prof. Matthias Scheffler (Fritz-Haber-Institute of the MPG)
  Paradigm shift by data-centric materials science jointly steered with and enabled by MPCDF

• Prof. Claudia Felser (Max Planck Institute for Chemical Physics of Solids)
  Topology and Chirality
  Topology, a mathematical concept, recently became a hot and truly transdisciplinary topic in condensed matter physics, solid state chemistry and materials science. All 200 000 inorganic materials were recently classified into trivial and topological materials: topological insulators, Dirac, Weyl and nodal-line semimetals, and topological metals [1]. The direct connection between real space: atoms, valence electrons, bonds and orbitals, and reciprocal space: bands and Fermi surfaces allows for a simple classification of topological materials in a single particle picture. More than 25% of all inorganic compounds host topological bands, which opens also an infinitive play-ground for chemistry [1,2]. Beyond Weyl and Dirac, new fermions can be identified in compounds that have linear and quadratic 3-, 6- and 8- band crossings that are stabilized by space group symmetries [3]. Crystals of chiral topological materials CoSi, AlPt and RhSi were investigated by angle resolved photoemission and show giant unusual helicoid Fermi arcs with topological charges (Chern numbers) of ±2 [4]. In agreement with the chiral crystal structure two different chiral surface states are observed. A quantized circular photogalvanic effect is theoretically possible in Weyl semimetals. However, in the multifold fermions with opposite chiralities where Weyl points can stay at different energies, a net topological charge can be generated. [5]. However, chirality is also of interest for chemists [6], especially because of the excellent catalytic performance of the new chiral Fermions AlPt and PdGa [7]. The open question is the interplay between Berry curvature, chirality, orbital moment and surface states.

• Prof. Moritz Helmstaedter (Max Planck Institute for Brain Research)
  Connectomics
  Brains are highly interconnected networks of millions to billions of neurons. For a century, we have not been able to map these connectivity networks at synaptic resolution. Only recently, using novel electron microscopy techniques and machine-learning based data analysis, the mapping of neuronal networks has become possible at a larger scale. This new field of connectomics is still limited by technology and requires efficient AI-based analysis of peta-to-exascale datasets, but it is already starting to provide exciting insights into how neuronal circuits operate in the brain. We are turning connectomics into a high-throughput screening technique for neuroscience, for discovering brain-implemented algorithms, which may inspire novel machine learning, to map the imprints of sensory experience onto neuronal networks in the brain, and to investigate connectome alterations in models of psychiatric disease.

• Prof. Volker Springel (Max Planck Institute for Astrophysics)
  The dark and luminous sides of cosmic structure formation
  Simulations of cosmic structure formation have come a long way. Nowadays, they are not only accurately predicting the dark matter backbone of the cosmic web and the internal structure of halos and their satellites far into the non-linear regime, but are also capable of following the baryonic sector with rapidly improving physical fidelity. I will review the methodology and selected successes of recent hydrodynamical galaxy formation simulations, and critically discuss some of the primary uncertainties in modelling strong, scale-dependent feedback processes. Finally, I discuss some of the challenges lying ahead in this field in the coming years.

• Prof. Hans-Joachim Bungartz (Technical University Munich)
  The new science - it's computational!
  Computers allow for computations, and computations revolutionize the sciences, which become computational. What is the essence of “computational”, what are its ingredients, what are its appearances? Is our ensemble of disciplines still appropriate, or do we rather need a new cartography of the scientific landscape, the interdisciplinary defining the new disciplines? What is the impact of all this on governance structures, on education? How do research infrastructures or computing centres have to adapt, to transform? The presentation wants to reflect a bit on these and other questions related to “computational”.

• Prof. Gerhard Hummer (Max Planck Institute of Biophysics)
  Computing in times of COVID
  We use molecular dynamics (MD) simulations to explore the molecular processes in SARS-CoV-2 infections, working closely with our experimental collaborators. MD simulations have given us a detailed understanding of the structure and motions of the spike proteins at the viral surface, their interactions with the host cell receptors and membranes, and the epitopes they present for antibody targeting. MD simulations also helped to reveal the molecular interactions allowing SARS-CoV-2 to interfere with the cellular immune defense and, in turn, to identify possible routes for therapeutic targeting. Overall, large-scale MD simulations help us to uncover some remarkable biology associated with viral infection and, as we hope, guide our fight against COVID-19.

Contact: office@mpcdf.mpg.de