Postdoc position: Oak Ridge National Labs Leadership Computing Facility (OCLF)
The Oak Ridge National Laboratory's Leadership Computing Facility (OLCF) is seeking a postdoctoral research associate for the project "Massively Parallel Block Structured Adaptive Mesh Refinement on Hybrid Architectures for Subsurface Flow Applications."
Predictive modeling of subsurface reactive flows is a daunting task because of the wide range of spatial and time scales involved. Uniform grids can resolve features only on the order of meters at best. The PFLOTRAN project is aimed at developing the tools and methods needed for the next-generation massively parallel codes for multiphase, multicomponent reactive flow and transport.
You will play a major role in developing scalable groundwater flow applications in PFLOTRAN with an emphasis on CO2 sequestration problems on massively parallel hybrid multicore-GPU architectures. This will involve research and development of efficient multilevel solution algorithms for PDE systems on dynamic block-structured adaptive grids. You will be part of a multidisciplinary team of scientists at national laboratories and universities, including geologists, applied mathematicians, and computer scientists.
Qualifications: The ideal applicant will have
* Ph.D. in Applied Mathematics, Computer Science, or related field
* Experience developing or using object-oriented block-structured adaptive mesh refinement libraries
* Experience developing or using multigrid and multilevel solution algorithms for elliptic PDE systems
* Knowledge of nonlinear solvers, Krylov solvers, and preconditioning techniques
* Programming experience in C++ and/or Fortran 90
* Excellent communication skills and fluency in English
* Experience with PETSc, Hypre, SAMRAI libraries
* Parallel programming experience with MPI
* Programming experience with CUDA is a plus
The position is for a period of two years.
To apply, please view the general OLCF computational scientist postdoctoral research associate listing at
and follow the instructions for application.