Research
Integrating quantum chemistry method development with performance-portable software for accelerator-heavy computing environments.
Core Research Themes
Multi-layer Adaptive Partitioning (MAP)
Extending MAP approaches to treat complex reactive environments while maintaining chemical accuracy.
- Designing hierarchical partitioning strategies that balance subsystem fidelity and computational cost
- Building reusable MAP analysis workflows for collaborators
Fragment-based Electronic Structure at Scale
Advancing interfragment coupling models for covalent systems.
- Rigorous treatments published in J. Phys. Chem. A and J. Chem. Phys.
- Modernizing GAMESS fragmentation modules for GPU-enabled supercomputers
- Maintaining libfrag as an open-source sandbox for new algorithms
High-Performance Computing & Accelerator Enablement
Optimizing quantum chemistry kernels for heterogeneous CPU/GPU architectures.
- MPI/OpenMP hybrids and CUDA implementations
- Benchmarking on DOE-class systems with Ames National Laboratory
- Ensuring novel methods land in production-quality code bases
Active Projects
Subsystem-local SAPT acceleration
Applying resolution-of-identity techniques to reduce the cost of exchange-repulsion terms.
Accelerated MAP workflow
Co-developing GPU-ready MAP components targeting realistic condensed-phase simulations.
GAMESS extreme-scale modernization
Deploying new fragmentation kernels across DOE leadership-class machines.
Collaborations & Partnerships
University of Colorado Denver
Guidez Lab & Lin Group: multiscale methods and computational chemistry applications
Ames National Laboratory
GAMESS development, exascale readiness, and novel architecture studies
MolSSI Community
Best practices for sustainable scientific software and mentorship
Interested in Collaboration?
Reach out to discuss research opportunities or learn more about these projects.
Get in Touch