An Implicitly Parallel Object-Oriented Matrix Library and its Application to Medical Physics

We introduce VLADYMIR, a matrix library that permits the development of array-based code in C++. It is especially useful for numerical simulation tasks and parallelises automatically, without any need for parallelisation-specific instructions. Thanks to the underlying data-parallel model, it shows up an excellent scalability, even on large parallel machines. VLADYMIR has been successfully tested so far on simulations of physical systems by cellular automata and lattice Boltzmann models. In this paper, an application to emboli detection in medical physics is presented. Occurrence of emboli in the brain presents a risk for cardio-vascular accidents, which one tries to avoid by ultrasonography methods. Simulations using lattice Boltzmann methods help to analyse the interaction of an embolus with ultrasounds and lead towards a better understanding of the structure of the signal in ultrasonography measurements. Key words. Data-parallelism, array-based computation, lattice Boltzmann models,...

• Type: Conference
• Year: 2003
• Where: IPPS
• Authors: Jonas Lätt, Bastien Chopard
• Keywords: Distributed And Parallel Computing, IPPS 2003, Lattice Boltzmann, Lattice Boltzmann Methods, Lattice Boltzmann Models