Mar 07

Print this Post

Meet the #GPUhackDD 2018 mentors

The GPU Hackathons would not be as successful, were it not for the mentors. Every one of our nine teams has been assigned two mentors who work closely with the teams to help them achieve their goals. And now – meet the GPU experts of the Dresden GPU Hackathon (in alphabetical order):


Dr. Ben Cumming is a computational scientist in the Software Software and Libraries group at the Swiss National Supercomputing Center.
He has extensive experience in developing, optimizing and porting scientific software for different HPC architectures.
Currently Ben is leading the development of Arbor, a performance portable library for simulation of large, distributed networks of multi-compartment neurons under the aegis of the HBP.
My name is Tobias Frust and I have studied Information System Technologies at TU Dresden. During my studies I was introduced to GPUs by a lecture dealing especially with CUDA and OpenACC. In line with my student research project and the diploma thesis at Helmholtz-Zentrum Dresden-Rossendorf (HZDR) I gathered practical experience in developing GPU applications to help solving scientific problems. After my studies, I joined the Computational Science Group at HZDR. I a am very much interested in new developments concerning GPU-accelerated applications and design concepts.
Andreas Herten is a post-doctoral researcher at Jülich Supercomputing Centre of Forschungszentrum Jülich. In the scope of the NVIDIA Application Lab and the POWER Acceleration and Design Centre, he studies and optimizes scientific applications to run efficiently on GPUs and GPU-equipped POWER systems. He teaches CUDA and OpenACC to scientists of different fields.
Jeffrey Kelling is a scientist in the computational science group at the Helmholtz-Zentrum Dresden-Rossendorf, concerned with high performance computing and deep learning applications in science. He obtained his diplom in statistical physics and wrote his Ph.D. thesis on massively parallel lattice Monte-Carlo simulations on GPUs.
Brent Leback is the Service and Support Manager for PGI. He has worked in various positions over the last 34 years in HPC customer support, math library development, applications engineering and consulting at QTC, Axian, PGI, STMicroelectronics, and NVIDIA.
My name is Alexander Matthes and I studied computer science with subsidiary subject biology at the Technical University Dresden with focus on visualization and GPU high performance computing. I finished my studies in May 2016. In September 2016 I started my Ph.D. in computer science at Technical University Dresden and Helmholtz-Zentrum Dresden – Rossendorf. I am working now on performance portability of our GPU accelerated plasma simulation and on the next step in parallel computing for parallel simulation software and large-scale data analysis.
Vishal Mehta is working as Developer Technology at NVIDIA in collaboration with activities around CSCS & ETH. Primary focus on GPU optimizations in HPC applications; weather models, fluid dynamics, combustion, bio-mechanics, signal processing etc.
He has experience in Linear Algebra, MPI, CUDA, OpenACC, scientific visualizations.
Julian Miller is a doctoral candidate at RWTH Aachen University and works as a research assistant in the HPC Group of the RWTH’s IT Center. He obtained his master’s degree in Computational Engineering Science at RWTH Aachen University in 2016. His research is centered on high-performance computing with a focus on development productivity and cost modeling for programming many- and multicore systems.
Guray Ozen works as a compiler engineer at NVIDIA corporation. I currently focus on compilers and runtimes for GPUs and heterogeneous systems, primarily targeting auto code generation techniques for GPU accelerators.
Before joining NVIDIA, I worked 4 years (from 2014 to 2018) as a PhD student researcher at Barcelona Supercomputing Center and I was funded by IBM deep learning center.
My work at BSC which was also part of my PhD research explored how to improve parallelization and optimizations for GPUs.
My key contribution was the MACC compiler which supports GPU offloading using OpenMP accelerator model directives.
Ludwig Schneider came in touch with GPU programming in 2012.
Applying CUDA in the context of soft matter simulations he received his Bachelor in physics in 2013 and his Master in 2015 from the University of Goettingen.
He is currently pursuing a Ph.D. in physics designing and working with accelerated HPC software using CUDA and OpenACC.
My name is Sebastian Starke and I am currently working in the Computational Science Department at the HZDR in Dresden. Before joining the HZDR I worked as an algorithm engineer in the field of automatic speech recognition. I studied mathematics at the OvGU in Magdeburg at a bachelors level before majoring in statistics in 2015. I am looking forward to supporting the teams mostly regarding algorithmic or math related questions and details.
Peter Steinbach has studied Physics at the University of Leipzig where he majored in Particle Physics with a Diploma Thesis at DESY Hamburg. He completed a PhD on data of the Large Hadron Collider (CERN, Switzerland) which he received in 2012. He then switched fields to become an IT specialist and Scientific Software Engineer at the Max Planck Institute of Molecular Cell Biology and Genetics (as a client of Scionics Computer Innovation GmbH) where he was introduced to GPUs and has been working with them ever since.
Jan Stephan is a Computer Science student at the TU Dresden with focus on HPC. In parallel to his studies he worked or interned in various HPC-related positions over the past years, including the Helmholtz-Zentrum Dresden-Rossendorf and NVIDIA.
He has experience with CUDA, OpenCL, OpenMP and vectorization as well as multi-GPU programming with MPI and NVSHMEM; his preferred programming language is modern C++.
Matthias Werner has studied Business Mathematics (B.Sc.) and Network Computing (M.Sc.) at the TU Freiberg, Saxony. During the study he learned CUDA and OpenGL through various research projects in Boolean Problems and material science. In the last two years he has been working for the GPU Center of Excellence at the TU Dresden in the fields of CUDA programming, GPU performance analysis and teaching. 2018 he joined the Computational Radiation Physics Group at the Helmholtz-Zentrum Dresden – Rossendorf.
My name is René Widera, I finished my education as IT specialist for application development in 2009. I lead the technical part of all software developemnt of the Computational Radiation Physics group at Helmholtz-Zentrum Dresden-Rossendorf. I supervise more than ten CUDA-related open source projects on GitHub, the most prominent being the world’s fastest 3D3V electromagnetic particle-in-cell code for plasma physics, PIConGPU. My special skills are template meta programming, asynchronous computing and low level optimization for many core architectures.

Permanent link to this article: https://gcoe-dresden.de/meet-the-gpuhackdd-2018-mentors/

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>