CompBioMed’s 19th e-seminar took place on 9 November 2021
Biomedical Supercomputing Applications
In the first part of the talk, the biomedical context is presented, while the second part is dedicated to parallelisation and performance issues and enhancement.
1st part. Organ level simulations represent both a challenge and an opportunity: the Virtual Patient modelling for precision medicine. It is the ultimate example on supercomputers usage for multiphysics / multiscale modelling problems. At organ level, modelling is done by tightly coupling different physics (e.g. fluid, tissue, electrophysiology, chemical reactions, heat, transport of large bodies, particles or species) with contributions from different time and space scales (cells, tissue, organ, system). To make things more complex, these problems present issues such as patient variability and comorbidities in complex geometries, with a extremely difficult validation. A way of facing these two issues is to run problems in a virtual patients population. All these things together make the use of supercomputers a decisive factor. In this talk we will define the challenges, and showcase them with some examples of the cardiovascular and respiratory systems.
2nd part. To make those multiphysics simulations possible and efficient, several algorithmic and computational optimizations have been carried. In particular, we will present classical actuations to enhance the strong and weak scalabilities of the code. These scalabilities are necessary to enable the road to exascale, but they should always be accompanied by efficiency measures. To this end, we have developed several original strategies that will be presented: a dynamic load balance strategy at node level; a co-execution strategy for heterogeneous architectures; a computationally enhanced Gauss-Seidel method for multiphysics coupling; an auto-tuning strategy to select the resources under efficiency criteria.
Presentations was given by Dr. Guillaume Houzeaux (Barcelona Supercomputing Centre) and Dr Mariano Vazquez (Barcelona Supercomputing Centre).
Guillaume studied physics at the Université de Montréal. After his bachelor studies, he carried out a Master of Applied Science in CFD at Concordia University. He then went to Spain to undertake a PhD in Domain Decomposition methods applied to CFD at UPC, Barcelona. In 2005, he had the opportunity to join the newly created Barcelona Supercomputing Center (BSC-CNS) to start the CFD and HPC research line in the Department of Computer Applications in Science and Engineering. At the present time, Guillaume is leading, together with other three team leaders, a group of 50 researchers whose main mission is to develop and apply the multi-physics parallel simulation platform ALYA. This code, part of the European PRACE benchmark suite, is an HPC-based software adapted to run efficiently in large-scale parallel computers. His research involves physical modelling, numerical algorithms and code development and optimization, all driven by the objective of efficient use of supercomputing resources. He has published around 80 papers in international journals and has been participating to numerous national and European projects (W2Plastics, PRACE, SHERLOC cleansky), centres of excellence (EoCoE, RAISE, CompBiomed), but also contributed to industrial contracts (JyD, Repsol, Iberdrola). He is also co-founder of ELEM Biotech, a BSC spinoff company.
Mariano is a Researcher at the Barcelona Supercomputing Center, where since 2005 he has co-led the Alya Project. He is also CSO/CTO of ELEM Biotech, a BSC spinoff company. ELEM’s mission is to speed up the technology transfer from BSC to the biomedical domain, putting our supercomputer-based tools in the hands of those that need them to improve healthcare. Alya is the in-house parallel multi-physics simulation tool, which currently involves around 50 researchers and developers. Alya is specifically designed to run efficiently in supercomputers, being capable of simulating problems of the greatest complexity. His main research lines fall within Computational Science, such as Computational Bio-Mechanics (particularly Solid Mechanics of organic tissue and Electrophysiology) at organ and system level. Following these lines, the team develops a simulation tool to study the cardiovascular and respiratory systems targeted to biomedical researchers in academia, medical devices sector and pharmaceutical industry. Infarction, ageing, aneurisms rupture risk, arrhythmias, stent design or drug delivery are among the topics where such a tool can become a decisive help.
This webinar series is run in collaboration with the VPH Institute