9th HiRSE Seminar

On December 01, 2022, 3.00pm, Adrian Kummerländer from KIT/IANM will continue the HiRSE Seminar with his talk on Lattice Boltzmann Performance Engineering in OpenLB - Refactoring a legacy code to state-of-the-art performance on heterogeneous HPC clusters.

Abstract: Lattice Boltzmann Methods (LBM) are an established mesoscopic approach for simulating a wide variety of transport phenomena. While they are uniquely suited to HPC usage, obtaining maximum performance across different targets and applications is an active topic of research.
The open source C++ framework OpenLB is one of the major freely available LBM research codes, being in active development since 2007. In recent years, a concerted and successful effort was undertaken to improve the performance on state-of-the-art heterogeneous HPC clusters by adding e.g. support for GPUs and vectorized execution on CPUs alongside general code improvements.
This talk will summarize the chosen refactoring approach alongside recent scalability benchmarks on the HoreKa supercomputer at KIT as well as some research results on optimizing single node GPU and SIMD CPU performance.

The talk will be held online. The link will be posted in the HiRSE_PS-internal HiRSE Mattermost, in the FZJ #rse Rocket.Chat channel, RSE mailing lists of the HiRSE partners (if any) and via the de-RSE mailing list and Mattermost. If you would like to attend, but think you may not get the link through these channels, please write to hirse@fz-juelich.de.

Location: virtual