BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:Europe/Stockholm X-LIC-LOCATION:Europe/Stockholm BEGIN:DAYLIGHT TZOFFSETFROM:+0100 TZOFFSETTO:+0200 TZNAME:CEST DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20230831T095754Z LOCATION:Davos DTSTART;TZID=Europe/Stockholm:20230627T100000 DTEND;TZID=Europe/Stockholm:20230627T103000 UID:submissions.pasc-conference.org_PASC23_sess110@linklings.com SUMMARY:Flash Poster Session - Part II DESCRIPTION:Poster\n\nThe aim of this rapid-fire session is to allow poste r presenters to introduce the topic of their poster and motivate the audie nce to visit them at the poster session scheduled on the following day. Pr esenters will have 30 seconds to engage the audience.\n\nThe Flash Poster Session is sponsored by the Royal Society of Chemistry Journals (https://w ww.rsc.org/). Winners of the poster competition will receive a voucher pri ze generously offered by the Digital Discovery (https://www.rsc.org/journa ls-books-databases/about-journals/digital-discovery/) and by PCCP (https:/ /pubs.rsc.org/en/journals/journalissues/cp#!recentarticles&adv) during the award session scheduled on the last day of the conference.\n\nP43 - Multi level and Domain-Decomposition Solution Strategies for Solving Large-Scale Phase-Field Fracture Problems\n\nThe phase-field approach for fracture pr opagation is a state-of-the-art technique for simulating crack initiation, propagation, and coalescence. In this approach, a damage field, called th e phase field, is introduced that characterizes the material state from in tact to fully broken. Even though the ...\n\n\nHardik Kothari (Università della Svizzera italiana); Alena Kopanicakova (Brown University, Università della Svizzera italiana); Patrick Zulian (Università della Svizzera itali ana, UniDistance Suisse); Maria Nestola (Università della Svizzera italian a); Edoardo Pezzulli and Thomas Driesner (ETH Zurich); and Rolf Krause (Un iversità della Svizzera italiana, UniDistance Suisse)\n------------------- --\nP38 - Loki v0.1.1: A Source-To-Source Translation Tool for Numerical W eather Prediction Codes and More\n\nAll known or presumed candidates for e xascale supercomputers will feature novel computing hardware or heterogene ous architectures, with GPUs currently being a cornerstone of this develop ment. Using these machines efficiently with today's numerical weather pred iction (NWP) codes requires adapting lar...\n\n\nMichael Staneker, Ahmad N awab, Balthasar Reuter, and Michael Lange (ECMWF)\n---------------------\n P56 - The P4est Software for Parallel AMR: A Shared Memory Workflow\n\nPar allel adaptive mesh refinement (AMR) is a key technique when simulations a re required to capture time-dependent and/or multiscale features. A forest of octrees is a data structure to represent the recursive adaptive refine ment of an initial, conforming coarse mesh of hexahedra. This poster prese n...\n\n\nMikhail Kirilin and Carsten Burstedde (University of Bonn)\n---- -----------------\nP33 - ICON-GPU for Numerical Weather Prediction – A Sta tus Report\n\nWeather prediction centers are always seeking ways to improv e the computational performance of their numerical weather prediction (NWP ) models while staying within budget. The era of ever improving scalar CPU has come to an end but massively multiprocessing GPUs are advertised as t he next step forwa...\n\n\nMarek Jacob (DWD); Dmitry Alexeev (NVIDIA Inc.) ; Daniel Hupp and Xavier Lapillonne (MeteoSwiss); and Florian Prill, Danie l Reinert, and Günther Zängl (DWD)\n---------------------\nP49 - Simulatin g Aquaplanet Using ICON with a GT4Py DSL Dynamical Core\n\nWe present the results of our efforts porting the dynamical core of the ICON climate and numerical weather prediction (NWP) model to GT4Py. GT4Py is a Domain Speci fic Language (DSL) designed for weather and climate applications, which al lows domain scientists to write performance portable climate an...\n\n\nCh ristoph Müller (MeteoSwiss); Abishek Gopal (ETH Zurich / CSCS); Nicoletta Farabullini (ETH Zurich); Till Ehrengruber (ETH Zurich / CSCS); Samuel Kel lerhals, Peter Kardos, and Magdalena Luz (ETH Zurich); Matthias Röthlin (M eteoSwiss); Enrique G. Paredes (ETH Zurich / CSCS); David Leutwyler and Be njamin Weber (MeteoSwiss); Rico Häuselmann and Felix Thaler (ETH Zurich / CSCS); Jonas Jucker (ETH Zurich); Linus Groner, Hannes Vogt, and Mauro Bia nco (ETH Zurich / CSCS); Anurag Dipankar (ETH Zurich); and Carlos Osuna an d Xavier Lapillonne (MeteoSwiss)\n---------------------\nP46 - Parallel Se cond Order Conservative Remapping on the Sphere\n\nWe present an MPI-paral lel implementation and analysis of a conservative second-order interpolati on method between arbitrary very high resolution spherical meshes supporte d by the Atlas library of ECMWF. Hence, meshes are those used by ECMWF’s I FS model: structured grids such as the reduced Ga...\n\n\nSlavko Brdar, Wi llem Deconinck, Pedro Maciel, and Michail Diamantakis (ECMWF)\n----------- ----------\nP53 - Ultra-High Resolution Simulations of Planetary Collision s\n\nGiant impacts (GI) form the last stage of planet formation and play a key role in determining many aspects like the final structure of planetar y systems and the masses and compositions of its constituents. A common ch oice for numerically solving the equations of motion is the Smoothed Parti cle Hydro...\n\n\nThomas Meier, Christian Reinhardt, Douglas Potter, and J oachim Stadel (University of Zurich)\n---------------------\nP50 - Towards a GPU-Enabled Linear-Response Algorithm in the SIRIUS Library\n\nElectron ic-structure approaches have become integral in materials science, physics and chemistry for studying existing and designing and discovering novel m aterials. Among the properties that can be studied, spectral properties of materials provide a wealth of information, and can be obtained from K...\ n\n\nGiannis D. Savva and Iurii Timrov (EPFL), Nicola Colonna (Paul Scherr er Institute), Anton Kozhevnikov (ETH Zurich / CSCS), and Nicola Marzari ( EPFL)\n---------------------\nP55 - Novel Geometric Deep Learning Surrogat e Framework for Non-Linear Finite Element Simulations\n\nConventional nume rical methods are computationally expensive in simulating non-linear pheno mena arising in mechanics. In this aspect, deep learning (DL) techniques a re being increasingly used for accelerating simulations in mechanics. Howe ver, existing DL methods perform inefficiently as the size an...\n\n\nSaur abh Deshpande (University of Luxembourg); Jakub Lengiewicz (University of Luxembourg, IPPT PAN); and Stéphane Bordas (University of Luxembourg)\n--- ------------------\nP37 - LIBRSB: Multicore Sparse Matrix Performance Acro ss Languages and Architectures\n\nLIBRSB (http://librsb.sf.net/) is a high ly interoperable multicore CPU-oriented library for sparse matrix computat ions.
It serves as a component in sparse linear solvers. LIBRSB build s upon its "RSB" hierarchical and coarse-grained sparse matrices storage. The RSB data structure and algorithms ...\n\n\nMichele Martone (Leibniz Su percomputing Centre)\n---------------------\nP52 - Tunable and Portable Ex treme-Scale Drug Discovery Platform at Exascale: the LIGATE Approach\n\nTo day digital revolution is having a dramatic impact on the pharmaceutical i ndustry and the entire healthcare system. The implementation of machine le arning, extreme-scale computer simulations, and big data analytics in the drug design and development process offers an excellent opportunity to low er...\n\n\nAndrea Beccari (Dompé farmaceutici), Silvano Coletti (Chelonia SA), Biagio Cosenza (Università di Salerno), Andrew Emerson (CINECA), Thom as Fahringer (University of Innsbruck), Daniele Gregori (E4 Engineering), Philipp Gschwandtner (UIBK), Erik Lindahl (KTH Royal Institute of Technolo gy), Jan Martinovic (IT4Innovations National Supercomputing Center), Gianl uca Palermo (Politecnico di Milano), and Torsten Schwede (University of Ba sel)\n---------------------\nP51 - Towards a Python-Based Performance-Port able Finite-Volume Dynamical Core for Numerical Weather Prediction\n\nWe p resent recent progress in the development of a high-performance Python imp lementation of the FVM non-hydrostatic dynamical core at ECMWF and its mem ber state partners. The FVM numerical formulation centred about 3D semi-im plicit time integration of the fully compressible equations with finite-vo ...\n\n\nStefano Ubbiali (ETH Zurich), Till Ehrengruber (ETH Zurich / CSCS ), Nicolai Krieger (ETH Zurich), Christian Kühnlein (ECMWF), and Lukas Pap ritz and Heini Wernli (ETH Zurich)\n---------------------\nP35 - Investiga ting the Mechanism of a Local Windstorm in the Swiss Alps Using Large-Eddy Simulations\n\nThe Laseyer windstorm is a local and strong wind phenomeno n in the narrow Schwende valley in northeastern Switzerland. The phenomeno n has raised the interest of meteorologists as it has - in the past - led to derailments of the local train. It is characterised by easterly to sout heasterly winds duri...\n\n\nNicolai Krieger (ETH Zurich), Christian Kühnl ein (ECMWF), and Michael Sprenger and Heini Wernli (ETH Zurich)\n--------- ------------\nP44 - Numerical Simulation of Gradual Compaction of Granular Materials and the Uncertainty Quantification of the Proposed Mathematical Model\n\nThe poster deals with mathematical modelling of granular materia ls and focuses on the process of their gradual compaction called ratchetti ng. The model of hypoplasticity introduced by E. Bauer et al. is investiga ted and the problem of stress-controlled hypoplasticity is considered. The behaviour of ...\n\n\nJudita Runcziková and Jan Chleboun (Czech Technical University in Prague)\n---------------------\nP32 - High-Throughput Compu tational Screening of Fast Li-Ion Conductors\n\nWe present a high-throughp ut computational screening to find fast Li-ion conductors to identify prom ising candidate materials for application in solid-state electrolytes. Sta rting with ~30,000 experimental structures sourced from COD, ICSD and MPDS repositories, we performed highly automated calcula...\n\n\nTushar Thakur , Loris Ercole, and Nicola Marzari (EPFL)\n---------------------\nP42 - Mu ltigrid in H(curl) on Hybrid Tetrahedral Grids\n\nIn many applications lar ge scale solvers for Maxwell's equations are an indispensable tool. This w ork presents theory and algorithms that are relevant to the solution of Ma xwell's equations as well as their implementation in HyTeG. We focus on mu ltigrid methods for the curl-curl-problem which arises...\n\n\nDaniel Baue r (Friedrich-Alexander-Universität Erlangen-Nürnberg)\n------------------- --\nP47 - Parallel Training of Deep Neural Networks\n\nDeep neural network s (DNNs) are used in a wide range of application areas and scientific fiel ds. The accuracy and the expressivity of the DNNs are tightly coupled to t he number of parameters of the network as well as the amount of data used for training. As a consequence, the networks and the amount...\n\n\nSamuel Cruz (Università della Svizzera italiana, UniDistance Suisse); Alena Kopa nicakova (Brown University, Università della Svizzera italiana); Hardik Ko thari (Università della Svizzera italiana); and Rolf Krause (Università de lla Svizzera italiana, UniDistance Suisse)\n---------------------\nP41 - M PI for Multi-Core, Multi Socket, and GPU Architectures: Optimised Shared M emory Allreduce\n\nIn the literature the benefits of shared memory collect ives especially allreduce have been shown. This intra-node communication i s not only necessary for single node communications but it is also a key c omponent of more complex inter-node communication algorithms [1]. In contr ast to [2], our impleme...\n\n\nAndreas Jocksch and Jean-Guillaume Piccina li (ETH Zurich / CSCS)\n---------------------\nP48 - ProtoX: A First Look\ n\nStencil operation is a key component in the numerical solution of parti al differential equations. Developers tend to use different libraries that provide these operations for them. One such library is Proto. It is a C++ based domain specific library designed to provide an intuitive interface that op...\n\n\nHet Mankad and Sanil Rao (Carnegie Mellon University), Phi l Colella and Brian Van Straalen (Lawrence Berkeley National Laboratory), and Franz Franchetti (Carnegie Mellon University)\n---------------------\n P30 - High Performance Computing Meets Approximate Bayesian Inference\n\nD espite the ongoing advancements in Bayesian computing, large-scale inferen ce tasks continue to pose a computational challenge that often requires a trade-off between accuracy and computation time. Combining solution strate gies from the field of high-performance computing with state-of-the-art st ati...\n\n\nLisa Gaedke-Merzhäuser (Università della Svizzera italiana), H aavard Rue (King Abdullah University of Science and Technology), and Olaf Schenk (Università della Svizzera italiana)\n---------------------\nP36 - Iterative Refinement With Hierarchical Low-Rank Preconditioners Using Mixe d Precision\n\nIt has been shown that the solution to a dense linear syste m can be accelerated by using mixed precision iterative refinement relying on approximate LU-factorization. While most recent work has focused on ob taining such a factorization at a reduced precision, we investigate an alt ernative via low-ra...\n\n\nThomas Spendlhofer and Rio Yokota (Tokyo Insti tute of Technology)\n---------------------\nP34 - Interpretable Compressio n of Fluid Flows Using Graph Neural Networks\n\nNeural network (NN) based reduced-order models (ROMs) via autoencoding have been shown to drasticall y accelerate traditional computational fluid dynamics (CFD) simulations fo r rapid design optimization and prediction of fluid flows. However, many r eal-world applications (e.g. hypersonic propulsion, ...\n\n\nShivam Barwey (Argonne National Laboratory) and Romit Maulik (Argonne National Laborato ry, University of Pennsylvania)\n---------------------\nP31 - High-Perform ance Computing by and for Patient Specific Mechanical Properties\n\nModeli ng the mechanical behavior of human trabecular bones improves technical ap plications and the treatment of fractures and bone or joint related diseas es. However, this type of bone consists of a large number of struts and pl ates, resulting in a highly anisotropic and patient specific behavior. F.. .\n\n\nJohannes Gebert, Ralf Schneider, and Michael Resch (High-Performanc e Computing Center Stuttgart, University of Stuttgart)\n------------------ ---\nP40 - Modeling a Novel Laser-Driven Electron Acceleration Scheme: Par ticle-In-Cell Simulations at the Exascale\n\nIntense femtosecond lasers fo cused on low-density gas jets can accelerate ultra-short electron bunches up to very high energies (from hundreds of MeV to several GeV) over a few millimeters or a few centimeters. However, conventional laser-driven elect ron acceleration schemes do not provide enough ch...\n\n\nLuca Fedeli (CEA ), Axel Huebl (Lawrence Berkeley National Laboratory), France Boillod-Cern eux and Thomas Clark (CEA), Kevin Gott (Lawrence Berkeley National Laborat ory), Conrad Hillairet (Arm), Stephan Jaure (Atos), Adrien Leblanc (ENSTA) , Rémi Lehe and Andrew Myers (Lawrence Berkeley National Laboratory), Chri stelle Piechurski (GENCI), Mitsuhisa Sato (RIKEN), Neil Zaim (CEA), Weiqun Zhang and Jean-Luc Vay (Lawrence Berkeley National Laboratory), and Henri Vincenti (CEA)\n---------------------\nP45 - Optimization of Non-Conventi onal Airfoils for Martian Rotorcraft with Direct Numerical Simulations Usi ng High-Performance Computing\n\nDesign of rotorcraft for Mars is challeng ing due to the very low density and low speed of sound compared to Earth. These conditions require Martian rotor blades to operate in a low-Reynolds -number (1,000 to 10,000 based on chord) compressible flow regime, atypica l of conventional, terrestrial helico...\n\n\nLidia Caros, Oliver Buxton, and Peter Vincent (Imperial College London)\n\nSession Chair: Jibonananda Sanyal (National Renewable Energy Laboratory) END:VEVENT END:VCALENDAR