Recent performance improvements to the DFT and TDDFT in GAMESS

Area of Science:

• Computational Chemistry
• Materials Science
• Quantum Mechanics

Background:

• The General Atomic and Molecular Electronic Structure System (GAMESS) is a widely used quantum chemistry package.
• Density Functional Theory (DFT) and Time-Dependent DFT (TDDFT) are crucial for molecular modeling.
• High-performance computing resources are extensively utilized for GAMESS calculations, demanding efficient solutions.

Purpose of the Study:

• To identify and implement performance improvements in the GAMESS software.
• To reduce the time-to-solution for DFT and TDDFT calculations.
• To enhance the efficiency of complex molecular system modeling.

Main Methods:

• Optimization of the exchange-correlation (XC) integration grid.
• Profiling and optimization of the core DFT code.
• Parallelization of TDDFT calculations.

Main Results:

• Substantial reductions in computational time for DFT and TDDFT methods.
• Improved performance metrics for the GAMESS quantum chemistry package.
• Successful implementation of identified software enhancements.

Conclusions:

• The implemented software enhancements lead to significant performance gains in GAMESS.
• These optimizations contribute to considerable savings in CPU hours for computational chemistry tasks.
• The improved GAMESS versions are readily available for the materials modeling community.