Dr. Fabian Nick
Principal Data Science; Mathematician
My name is Fabian Nick and you are probably here because you read my CV, so thank you for that. I really appreciate the time you are taking to get to know me. This page just provides some additional content to accompany the information in my CV.
I am enthusiastic about the interface between applied mathematics, computer science and business. For me, work has to coincide with my passion and I love driving projects with my colleagues and clients.
Horn & Company Data Analytics GmbH (since 2019)
In my role as a Data Science Consultant and Manager, I engage in Data Science projects spanning various clients and industries, primarily within the German landscape. Within these endeavors, I address the complexities associated with data acquisition, storage, and utilization. My approach encompasses a holistic fusion of management consulting and Data Science expertise within a unified project framework, aimed at optimizing value delivery to our clients.
Publications:
Fraunhofer SCAI (2013 - 2019)
SAMG
SAMG is a library of scalable algorithms for the solution of large, sparse linear systems of equations which is used in many industrial applications. It supports both the MPI and the OpenMP parallelisation model as well as a hybrid version. My focus is primarily on the customers from the groundwater simulation industry which means that I provide consulting for those customers regarding both technical as well as numerical issues. In addition to that, I am part of the general SAMG development as well.
A short talk on SAMG in HPC (starts around 04:51)
Publications:
PLUM, Hans-Joachim, et al. Parallel algebraic multigrid. In: Scientific Computing and Algorithms in Industrial Simulations. Springer, Cham, 2017. S. 121-134.
MESHFREE
Together with the Fraunhofer ITWM I was part of the development of a new product called “MESHFREE” for which I adapted the SAMG library to suit the needs of the “Finite Pointset Method”, a method which is used to simulate processes occurring in fluid and continuum mechanics. Unlike classical CFD methods, MESHFREE does not use a grid to discretize the simulation domain, but uses a point cloud instead. This makes the method very suitable for simulations with free surfaces or moving geometries. MESHFREE is a highly parallel MPI-based code that is used on up to 512 CPU cores in industrial environments.
The official MESHFREE website can be found at www.meshfree.eu.
Publications:
NICK, Fabian. Algebraic Multigrid for Meshfree Methods. Thesis, University of Bonn. 2020.
METSCH, Bram; NICK, Fabian; KUHNERT, Jörg. Algebraic multigrid for the finite pointset method. In: Comput. Visual Sci. 23, 3 (2020). https://doi.org/10.1007/s00791-020-00324-3
NICK, Fabian; METSCH, Bram; PLUM, Hans-Joachim. Linear Solvers for the Finite Pointset Method. In: International Conference on Computational Engineering. Springer, Cham, 2017. pp. 89-110.
NICK, Fabian; PLUM, Hans-Joachim; KUHNERT, Jörg. Parallel Detection of Subsystems in Linear Systems Arising in the MESHFREE Finite Pointset Method. In: International Workshop on Meshfree Methods for Partial Differential Equations- Springer, Cham, 2017, pp. 93-115