Lab Introduction

The Fujio Cho Engineering Technology is a new department in the Pigman College of Engineering. I am special title faculty with a 20% research component. My background is in computational electromagnetic and scientific software engineering. My main research interest, which dovetails with the mission of the Engineering Technology Department, is the application of best software engineering practices like DevOps, Continuous Integration, Continuous Delivery, Reproducibility, Portability, and Traceability to the development process of scientific software in general and high-performance computing in particular.

Multi-scale data driven modeling of radiative transport through thermal protection systems

Co-PI, PI: Savio Poovathingal, Mechanical Engineering

Students

None yet from ET (ET does not have a graduate program)

Modeling Transitional and Turbulent Flows with Surface Ablation

Co-PI Jens Hannemann, PI: Alexandre Martin

Students

None yet from ET (ET does not have a graduate program)

Software

HPC projects typically involve a C++ or Fortran 90 code base with several dependencies, such as HDF5, Eigen, PETSc, MPI (in various implementations), ParMETIS, Mutation++, and others. Managing these dependencies in a manner that allows research group software to be built out of the box is challenging, and the traditional HPC modules approach is not flexible enough to handle all possible combinations of libraries and compilers, especially in the case of the non-standardized but relatively complex ABI of the C++ language. We actively research containerization as well as build and test automation and the impact of different machine-, language-, and cluster architectures on the reproducibility of test and production computation results. A promising path forward seem to be semi-open container formats specialized for HPC, such as Apptainer and Singularity, which apply to the existing code bases on LCC and MCC.
C++ and Fortran Compilers (GNU/Clang/Intel), CMake, Boost, Qt, Apptainer, Jenkins, Git, Gitolite, Tau (Tuning and Analysis Utilities), MPI (OpenMPI, IntelMPI, MPIch), OpenMP, CUDA

Publications

Grants

A. Martin (PI), S. Bailey, J. Hannemann: Modeling of High-Speed Transitional and Turbulent Flows over Ablative Surfaces - NASA

S. Poovathingal (PI), H. Poonawala, J. Maddox, J. Renfro, J. Hannemann: Multi–scale data driven modeling of radiative transport through thermal protection systems - NASA EPSCoR

Browser not supported