Biography
Andy Nonaka is a Staff Scientist and Group Lead of the Center for Computational Sciences and Engineering at Lawrence Berkeley National Laboratory. His research interests include high-performance implementations of multiphysics and multiscale algorithms for partial differential equations, with a recent emphasis on electrodynamical, fluid dynamical, and material science applications. More generally, he is interested in adaptive structured mesh algorithms, particle and particle/mesh methods, linear solvers, and machine learning techniques to accelerate scientific discovery.
SRP Project Title
High-Performance Multiphysics and Multiscale Modeling with AMReX
Topical Areas
Applied Computer Science; Applied Mathematics; Atmospheric Sciences; Computer Science; Electrical, Electronic, and Information Engineering; Fluid and Plasma Physics; High Performance Computing; Open Source Software; Performance Evaluation and Benchmarking; Software Engineering
Abstract
The Center for Computational Sciences and Engineering (CCSE) at LBNL is a high-performance modeling and simulation group interested in applications described by complex partial differential equations including fluids (industrial, biological, and micro/nanoscale), plasma physics, microelectronics, and environmental flows. The underlying software framework we develop and use is called AMReX, a GPU/exascale-enabled block-structured adaptive mesh refinement software framework that also includes particle/mesh, embedded boundary, and linear solver capabilities. We also develop data-driven / machine-learning enhancement, and code coupling strategies to both accelerate and increase the fidelity of our codes. We are looking for collaborators with common interests in the computer science, applied mathematics, and/or application science aspects of research performed within CCSE.
Desired Skills
C++, python, numerical methods, partial differential equations, domain knowledge in fluids, plasmas, microelectronics, and/or environmental flow.
Lightning Talk Title
High-Performance Multiphysics and Multiscale Modeling with AMReX
Keywords
partial differential equations, structured mesh, particle/mesh, data/ML-driven model acceleration, industrial fluids, biofluids, micro/nanoscale fluids, plasmas, microelectronics, environmental flow.