GPU-Accelerated Direct Numerical Simulations with an Immersed Boundary Method
Please login to view abstract download link
The rapid progress of technology in recent years has paved the way for Graphical Processing Units (GPUs) to become a prominent player in the world of high-performance computing. With their highly parallel architecture and fast processing capabilities, GPUs have become a very valuable tool for researchers and engineers in the field of fluid mechanics. In this work, we exploit conveniently the architecture of these devices, presenting a novel parallelization for running direct numerical simulations massively with a significant performance boost.