NASA uses parallel programming directives to model the Space Shuttle Launch Vehicle
Our mission is to perform cutting-edge research and development in programming languages, compilers, machine learning, tools and applications for performance, productivity and energy efficiency on High Performance Computing and Big Data platforms.
In our research, we strive to develop approaches and novel implementation techniques that anticipate architectural changes and emerging user needs; enhance application performance; facilitate programmer productivity; and provide implementations of our ideas.
We are active members of open standards such as OpenMP, OpenACC, the MultiCore Association (MCA), Unified Communication X (UCX) and OpenSHMEM. Our contributions to the HPC community include: the reference implementation for OpenSHMEM; the open-source OpenUH compiler; LLVM enhancements.
Please see current research positions & project opportunities.
The group is led by Dr. Barbara Chapman.
The research interests of Exascallab broadly span several areas of High Performance Computing. Some of our topics of interest include:
- Compiler Development and Optimizations for HPC, Big Data and Machine Learning
- We actively contribute to the LLVM compiler framework
- OpenMP programming model for Exascale systems
- Partitioned Global Address Space (PGAS) programming model
- Fault Tolerance techniques for Large Scale HPC systems
- Energy efficiency in HPC
