Postdoc: CFD/HPC Algorithms & Data-Driven ML

Southwest Virginia Higher Education Center; Charlottesville (VA); On-site; USD 48,000 - 90,000; Full time

Job summary

The Computational Fluid Dynamics (CFD) and Propulsion Laboratory at the University of Virginia invites applications for a Postdoctoral Research Associate position. The role involves developing advanced numerical algorithms for large-scale CFD on HPC platforms and contributing to data-driven and physics-driven learning methods.

The successful candidate will implement, test, and analyze scalable solvers and domain decomposition techniques, with a focus on HPC performance and mathematical rigor.

Qualifications

• PhD required by start date; within 5 years post-PhD.
• Research experience in CFD methods, numerical methods and analysis, stability analysis, scalable solvers and domain decomposition methods.
• Proficiency in C++ programming, parallel programming, and GPU programming.
• Exposure to stochastic PDEs, stochastic processes, model reduction, surrogate model development, and uncertainty quantification.

Responsibilities

• Develop, implement, and analyze efficient numerical methods for next-generation CFD on HPC platforms.
• Act as an algorithm developer within the CFD and Propulsion Laboratory; contribute to scalable solvers and learning techniques.

Skills

CFD methods

Numerical methods and analysis

Stability analysis

Scalable solvers

Domain decomposition methods

C++ programming

Parallel programming

GPU programming

Stochastic PDEs

Stochastic processes

Model reduction

Surrogate model development

Uncertainty quantification

AI/ML

Supercomputing

Large-scale computational infra

Education

PhD in Mathematics or Statistics or related field

Job description

The Computational Fluid Dynamics (CFD) and Propulsion Laboratory at the University of Virginia invites applications for a Postdoctoral Research Associate position. The role involves developing advanced numerical algorithms for large-scale CFD on HPC platforms and contributing to data-driven and physics-driven learning methods.

The successful candidate will implement, test, and analyze scalable solvers and domain decomposition techniques, with a focus on HPC performance and mathematical rigor.