David Trebotich

David Trebotich is an expert in computational fluid dynamics (CFD) and high performance computing (HPC). His research has involved end-to-end conception, development and analysis of advanced algorithms for scalable, high resolution simulation of complex flows in multiscale, multiphysics systems. He is the original developer of the suite of application codes, known as Chombo-Crunch, which simulates flow, transport and reactions in heterogeneous media. The Chombo-Crunch capability is based on a novel adaptive embedded boundary algorithm developed by David for solving flow and transport PDEs in arbitrarily complex geometry. More recently, he has focused on pore scale applications with multiphysics coupling on GPUs; these range from geological (subsurface) to engineering and manufacturing (energy storage). David’s simulations have achieved unprecedented scale and resolution on DOE supercomputers for flow and transport problems in complex, heterogeneous systems including a CFD simulation of nearly one half trillion degrees of freedom which is the largest CFD simulation in complex geometry to date. He is one the largest users, if not the largest, in the DOE complex of supercomputers having been awarded the equivalent of nearly 1.5 billion computational hours over the past 12 years. His current research goals include combining AI/ML-driven decision making with PDE-based simulation to develop stochastic AMR for adaptive modeling, as well as new application code development in energy storage and manufacturing. David has been a staff scientist in the Applied Numerical Algorithms Group at Lawrence Berkeley National Laboratory since 2009. Prior to the Berkeley lab, David was staff scientist at the Lawrence Livermore National Laboratory in the Center for Applied Scientific Computing from 2001 to 2009, and before that a post-doc at UC Berkeley after completing his Ph.D. there in 1998.