Speaker: Melven Röhrig-Zöllner, German Aerospace Center (DLR)

Title: Performance of linear solvers in tensor-train format on current multicore architectures

Date and time: Tuesday, February 27, 2024, 2:00-3:00 p.m. CET

Slides: https://hpc.fau.de/files/2024/03/NHR_PerfLab_2024-02-27_Performance-of-linear-solvers-in-TT-format.pdf

Abstract:

This talk discusses the node-level performance of numerical algorithms for handling high-dimensional problems in a compressed tensor format.

Short Bio:

Melven Röhrig-Zöllner studied Computational Engineering Science at the RWTH Aachen. Since 2014, he works as a researcher in the HPC department of the Institute for Software Technology of the German Aerospace Center (DLR). His research focuses on the performance of numerical methods, in particular in the field of numerical linear algebra. He also supports scientific software development for HPC systems in the DLR, e.g., concerning software engineering practices and testing parallel codes.