Julia for Scientific Computing and Data Science
Information
The Julia programming language [1] represents a modern approach to scientific computing and data science, combining a user-friendly, expressive syntax similar to Python with the high computational performance found in C/C++ and Fortran. This dual advantage makes Julia a compelling choice for a wide range of computational tasks, from quickly prototyping innovative ideas to running complex, large-scale production workloads on diverse hardware platforms.
Our event seeks to connect users and developers at the University of Augsburg, whether they are new to using Julia for scientific research or already have some experience. We aim to identify shared challenges that could benefit from joint efforts and to explore potential future Julia-focused collaborations. The event will include a series of Julia presentations to inspire dialogue, including a brief introduction to performance optimization, alongside an open-topic birds-of-a-feather session for informal discussions.
Organizers: Valentin Churavy, Michael Schlottke-Lakemper, Tatjana Stykel
Date: Wednesday, 18th December 2024
Registration
Registration is not mandatory but appreciated, since it allows us to gather information about people interested in using Julia for computational science. Thus, please consider registering even if you are not able to attend the event but would still like to be kept in the loop about potential future activities.
Schedule
Schedule
10:00 |
Welcome |
10:05 |
Performance optimization with Julia: a short introduction Valentin is a long-term core developer of the Julia language and an expert on using Julia for high-performance computing. He will give an overview of how to achieve high performance with Julia in general, and demonstrate practical techniques for analyzing and enhancing the computational efficiency of Julia code on CPUs. |
10:45 |
Lightning talks: Julia in computational science Niklas Neher: TrixiParticles.jl: Particle-based multiphysics simulations Simon Candelaresi: Adaptive model selection in coupled multiphysics simulations with Trixi.jl Martin Hermann: Ferrite.jl as a Finite Element Library in Julia Jonas Püschel: Implementing custom quantum chemistry solvers in DFTK.jl The lightning talks will provide insights into several Julia-based research projects and highlight benefits – and challenges – of using Julia for computational science. |
11:25 |
Birds of a Feather |
12:15 |
Closing remarks & next steps |
afterwards |
Joint lunch at mensa (optional) |
Directions
University of Augsburg
Building I2, Room 1309Universitätsstraße 12a
86159 Augsburg
Germany