6th European Seminar on Computing

June 3 - 8, 2018
Pilsen, Czech Republic

List of Minisymposia

If you are interested in organizing a minisymposium, please send a tentative title and a brief description to conference(at)esco2018(dot)femhub(dot)com.

Particle Simulations on GPUs

Nicolin Govender (govender.nicolin@gmail.com, University of Surrey, Guildford, United Kingdom and Research Center Pharmaceutical Engineering Graz, Austria), Daniel Wilke (University of Pretoria, South Africa), Charley Wu (University of Surrey, Guildford, United Kingdom).

In this mini-symposium we will explore the impact of the GPU in engineering simulations and glimpse into the future of simulations and computational engineering. Industrial-scale simulations using particle-based methods, such as the discrete element method, remain a big challenge, but the GPU architecture is changing that perception fast, as demonstrated by the recent advancement of the open-source framework Blaze-DEM. However, engineering simulation is still characterized by either the analyze-wait-modify-analyze cycle or more recently the batch analyze-wait-modify-batch analyze cycle. The GPU is enabling a new paradigm denoted to interactive simulation and design (ISD), allowing engineers to simulate process changes before being actually implemented.

While this mini-symposium is primarily focused on the algorithms and computational implementation of particle based methods using DEM, SPH, LBM on the GPU, applications of these methods in science and engineering will also be explored.

Computers and ICT in Mathematics Education

Gabriel Aguilera-Venegas, (gabri@ctima.uma.es, Universidad de Málaga, Spain), José Luis Galán-García (jl_galan@uma.es, Universidad de Málaga, Spain), Eugenio Roanes-Lozano (eroanes@mat.ucm.es, Universidad Complutense de Madrid, Spain), Pavel Solin (pavel@nclab.com, University of Nevada, Reno, USA)

The increasing use of computers and ICT (Information and Communication Technologies) in every kind of activity (industrial, academic, social, ...), is nowadays a fact that must be addressed. Specifically in Education, the computer and ICT are being used from different point of views in order to develop different Education strategies and techniques (programming, e-learning, blended learning, open and distance learning, learner-centered environments, …). It is very important to know the new trends in the use of Computer and ICT in Education since it is a field in constant evolution. In this minisymposium, proposals dealing with the use of Computers and ICT in Mathematics Education are welcome. The minisymposium will promote the outreach of new experiences, application of new educational models and techniques in Mathematics Education in which the use of computers and ICT have an key role.

Papers from talks presented at this minisymposium can be submitted to a special issue of The International Journal for Technology in Mathematics Education (IJTME, ISSN: 1744- 2710). IJTME is indexed in SCOPUS and ESCI (Emerging Sources Citation Index of The Web of Science), among other databases.

Smart Applications of Scientific Computing

Gabriel Aguilera-Venegas, (gabri@ctima.uma.es, Universidad de Málaga, Spain), José Luis Galán-García (jl_galan@uma.es, Universidad de Málaga, Spain), Antonio Hernando (ahernando@etsisi.upm.es, Universidad Politécnica de Madrid, Spain), Eugenio Roanes-Lozano (eroanes@mat.ucm.es, Universidad Complutense de Madrid, Spain)

Nowadays there is a wide variety of mathematical software available: computer algebra systems, technical computing languages, automated deduction systems, ... This minisymposium is devoted to practical real-world applications of this software in fields like: transportation engineering, electrical engineering, medicine, knowledge based systems, smart cities, accelerated time simulations, models of queuing systems, ... (this is not an exhaustive list). The focus will be on advanced and smart applications with a nontrivial mathematical background.

Computational Methodologies for Next-Generation Climate Models

Xylar Asay-Davis (Los Alamos National Laboratory, USA), Joseph H. Kennedy (Oak Ridge National Laboratory, USA), Salil Mahajan (Oak Ridge National Laboratory, USA), Irina Tezaur (ikalash@sandia.gov, Sandia National Laboratories, USA)

The development and application of global climate models for understanding and predicting the effects of global climate change and sea-level rise is critical, since it can direct energy and infrastructure planning, as well as inform public policy. Earth System Models (ESMs), which are global climate models including biogeochemistry, integrate the interactions between atmosphere, ocean, land, ice, and biosphere to enable the simulation of the state of regional and global climate under a wide variety of conditions. In recent years, there has been a push to develop “next generation” ESMs, models which: (1) are able to perform realistic, high-resolution, continental scale simulations, (2) are robust, efficient and scalable on next-generation hybrid systems (multi-core, many-core, GPU, Intel Xeon Phi) towards achieving exascale performance, and (3) possess built-in advanced analysis capabilities (e.g., sensitivity analysis, optimization, uncertainty quantification).

This minisymposium will consist of talks describing new and ongoing research in the development of accurate and tractable “next-generation” models for stand-alone climate components (e.g., atmosphere, land-ice, sea-ice, ocean, land, biogeochemistry), as well talks addressing the challenges in coupling climate components for integration into ESMs. Of particular interest are:

  1. efficient computational strategies and software for tackling the complex, nonlinear, multi-scale, multi-physics problems arising in climate modeling, with an eye towards next-generation hybrid platforms, and
  2. advanced analysis techniques that can inform/enhance existing models through the incorporation of observational data, e.g., approaches for model initialization/calibration, uncertainty quantification and data assimilation.

Discontinuous Galerkin methods and their applications

Stefano Giani (stefano.giani@durham.ac.uk, Durham University, United Kingdom)

Discontinuous Galerkin (DG) methods are finding more and more applications every day. DG methods were originally introduced for solving hyperbolic equations, now they have been adopted for any kind of equation linear, non-linear, steady, for steady-state and time-dependent problems and for systems of PDEs also. The success of DG methods is because of their unique characteristics like the freedom in constructing meshes or the lack of continuity across faces. The aim of this minisymposium is to present a broad survey of recent work on discontinuous Galerkin methods and their applications. Expected topics of discussion include but are not limited to fluid dynamics, mechanics, electromagnetism, multiphysics problems, topology optimisation, fracture propagation, stochastic DG methods, nonlinear problems, mesh adaptivity, polygonal DG elements.

Anisotropic mesh adaptation in scientific computing

Vit Dolejsi (dolejsi@karlin.mff.cuni.cz, Charles University, Czechia)

Anisotropic mesh adaptation exhibits an efficient tool for the numerical solution of various type of partial differential equations with applications in scientific computing. It can significantly reduce the number of degrees of freedom and the computational time necessary to achieve the required accuracy. This minisymposium is devoted to the development and new progress of anisotropic mesh adaptation techniques.

Linear Algebra on GPUs

Jonas Thies (Jonas.Thies@dlr.de, DLR, Germany), Dominik Ernst (Erlangen Regional Computing Center, Germany), Peter Zaspel (University of Basel, Switzerland)

At the heart of many applications lie linear algebra tasks such as solving large sparse or dense systems of equations or eigenvalue problems. These costly core problems may consume most of the computing time of a simulation, and their efficient parallel solution remains a large field of research.

On the hardware side, GPUs continue to excel in terms of flop rates and memory bandwidth, making them attractive for both compute intensive (dense) and memory-bounded (sparse) linear algebra.

This minisymposium aims to give a glimpse at the state of affairs, from suitable algorithms to available software and applications already using GPU-accelerated linear algebra. Particular topics of interest are abstraction layers to make GPUs accessible to a broad user base, software for heterogeneous (e.g. CPU/GPU) platforms, performance engineering and experience in real-world applications.

Computational Statistics

Anna Panorska (ania@unr.com, University of Nevada, Reno, USA)

The “Big Data” data sets needed to be analyzed and converted to information come from many areas including engineering, business, biology-genomics, medicine, climate sciences and weather. The data comes in many different forms such as numbers, words, sequences, sentences, functions, or images. The modern analysis methods have to address both the statistical and computational issues in order to be effective. This minisymposium will explore the many different areas of statistics and data science where computation plays an important role.

Software Workshops

If you are interested in organizing a software workshop, please send a tentative title and a brief description to conference(at)esco2018(dot)femhub(dot)com.

preCICE - A Coupling Library for Partitioned Multi-Physics Simulations

Organizer: Benjamin Uekermann (uekerman@in.tum.de, Technical University of Munich, Germany)

preCICE (Precise Code Interaction Coupling Environment) is a coupling library for partitioned multi-physics simulations, including, but not restricted to fluid-structure interaction and conjugate heat transfer simulations. Partitioned means that preCICE couples existing programs (solvers) capable of simulating a subpart of the complete physics involved in a simulation. This allows for the high flexibility that is needed to keep a decent time-to-solution for complex multi-physics scenarios. The software offers methods for transient equation coupling, communication means, and data mapping schemes. Ready-to-use adapters for well-known commercial and open-source solvers, such as OpenFOAM, SU2, or CalculiX, are available. Adapters for in-house codes can be implemented and validated in only a few weeks. preCICE is an open-source software under the LGPL3 license. github.com/precice/precice.

MercuryDPM - A Fast, Flexible and Accurate particle solver

Organizers: Deepak Tunuguntla, Thomas Weinhart, and Anthony Thornton, (d.r.tunuguntla@utwente.nl, t.weinhart@utwente.nl, a.r.thornton@utwente.nl, all organizers are from University of Twente, Netherlands)

A variety of granular materials in nature, academia and industry are often subjected to external forces. These include shaking, shearing and many other forces one could imagine of. Although experiments provide good insights in a granular process, simulations are inevitable as they play a crucial role in thoroughly understanding the dynamics of any granular process. As a result, in 2009, Anthony Thornton and Thomas Weinhart started MercuryDPM - a fast, flexible and accurate particle solver - at the University of Twente in The Netherlands.

MercuryDPM is a fully open-source particle simulation tool written in C++11, with a worldwide developer and user-base. It contains a large range of contact models, allowing for simulations of complex interactions such as sintering, breakage, plastic deformation, wet-materials and adhesion, all of which have important industrial applications. The code also contains novel complex wall generation techniques, that can exactly model real industrial geometries. Additionally, MercuryDPMs’ state-of-the-art built-in statistics package constructs accurate three-dimensional continuum fields such as density, velocity, structure and stress tensors, providing information often not available from scaled-down model experiments or pilot plants. The statistics package was initially developed to analyse granular mixtures flowing over inclined channels, and has since been extended to investigate several other granular applications. In this workshop, we will introduce the audience to MercuryDPM and briefly illustrate these novel techniques. mercurydpm.org.

pyMOR - Model Order Reduction with Python

Organizer: René Milk (rene.milk@wwu.de, Applied Mathematics Münster, Germany)

Reduced basis methods are projection-based model order reduction techniques for reducing the computational complexity of solving parametrized partial differential equation problems. Written in Python, pyMOR is a freely available, open source software (BSD-2 licensed), library of model order reduction algorithms. As its main design feature, all reduction algorithms in pyMOR are implemented generically via operations on well-defined vector array, operator and discretization interface classes. This allows for an easy integration with existing open-source high-performance partial differential equation solvers without adding any model reduction specific code to these solvers. Currently, there are bindings available for FEniCS, NGSolve, deal.II, and dune-gdt.

Please check pymor.org/esco18/ for details about this hands-on session.


Back to top

Copyright © 2013-2018 FEMhub All Rights Reserved.

FEMhub and the ESCO conferences are not affiliate programs of the University of Nevada, Reno.