Research Scientist - CFD on Exascale Architectures

The National Institute of Aerospace, located in Hampton, Virginia, is seeking candidates for a Research Scientist to conduct research and development aimed at exascale-class computational fluid dynamics simulations. The analysis and design of aerospace vehicles requires high-fidelity simulations capturing broad ranges of temporal and spatial scales previously considered intractable with the current class of petascale computing systems. A successful candidate is expected to develop, implement, and demonstrate computationally-performant implementations of existing and nascent algorithms at scale, targeting GPU-based High-Performance Computing (HPC) systems while retaining performance portability to CPU-based architectures.

 

The successful candidate will work as part of a tightly integrated interdisciplinary team of researchers at NIA, NASA, and in academia to advance the capabilities of interest.  Results of this work are expected to be published in both conference proceedings and journal papers.

·        Develop, implement, and demonstrate optimal performance of computational kernels critical to exascale-class simulations of aeronautics and space applications relevant to NASA missions

·        Collaborate within a multidisciplinary team, leveraging GPU-based HPC systems and state of the art programming models to realize a sustainable, performance-portable computational toolset for next-generation CFD simulations

 

·        MS in engineering, applied mathematics, computer science, physics, or equivalent, with strong background in HPC and computational science for large-scale applications.

·        Demonstrated experience in GPU programming techniques for real-world applications; ability to achieve and demonstrate optimal performance for both memory and compute-bound motifs.

·        A desire to quickly learn and apply new paradigms relevant to current and emerging architectures, and an eagerness to work in a team environment.

·        Ability to successfully complete a background investigation required to obtain a NASA badge.

·        US citizenship or permanent resident status is a bona fide requirement.

 

·        PhD preferred.

·        Previous experience targeting performance portability and the use of programming abstraction models

·        Previous experience working with unstructured grids and implicit solution techniques

·        Knowledge of fluid mechanics, numerical methods, and/or physical modeling techniques