Human-Scale Blood Flow Simulation Using Intel® OSPRay Studio
HemeLB is a highly scalable 3D blood flow solver capable of generating high-resolution simulations of blood flow through human-scale vasculatures. Postprocessing such simulations is a significant challenge due to the volume of data generated. We use Intel® OSPRay Studio to visualize the generated data time series directly on the production machine—SuperMUC-NG at Leibniz-Rechenzentrum (LRZ). We use a custom input plug-in to efficiently map the simulation data while eliminating complex preprocessing. This allows the users to quickly observe the full domain simulated by HemeLB for assessment.
Elisabeth Mayer has been working at the Leibniz Supercomputing Centre since 2017. Her research fields include virtual reality, rendering workflows, and visualization of complex datasets. She focuses on the development of a rendering workflow using high-performance computing resources to create 360° 3D videos.
Jon McCullough joined the University College London in 2019 to continue the development of the HemeLB blood flow simulation code. This work has particularly focused on developing a self-coupled version of the code for simultaneous human-scale arterial and venous flow modeling.