The General Atomic and Molecular Electronic Structure System (GAMESS): Novel Methods on Novel Architectures.

The primary focus of GAMESS over the last 5 years has been the development of new high-performance codes that are able to take effective and efficient advantage of the most advanced computer architectures, both CPU and accelerators. These efforts include employing density fitting and fragmentation methods to reduce the high scaling of well-correlated (e.g., coupled-cluster) methods as well as developing novel codes that can take optimal advantage of graphical processing units and other modern accelerators. Because accurate wave functions can be very complex, an important new functionality in GAMESS is the quasi-atomic orbital analysis, an unbiased approach to the understanding of covalent bonds embedded in the wave function. Best practices for the maintenance and distribution of GAMESS are also discussed.

Recent developments in the general atomic and molecular electronic structure system.

A discussion of many of the recently implemented features of GAMESS (General Atomic and Molecular Electronic Structure System) and LibCChem (the C++ CPU/GPU library associated with GAMESS) is presented. These features include fragmentation methods such as the fragment molecular orbital, effective fragment potential and effective fragment molecular orbital methods, hybrid MPI/OpenMP approaches to Hartree-Fock, and resolution of the identity second order perturbation theory. Many new coupled cluster theory methods have been implemented in GAMESS, as have multiple levels of density functional/tight binding theory. The role of accelerators, especially graphical processing units, is discussed in the context of the new features of LibCChem, as it is the associated problem of power consumption as the power of computers increases dramatically. The process by which a complex program suite such as GAMESS is maintained and developed is considered. Future developments are briefly summarized.

Improving efficiency of semi-direct møller-plesset second-order perturbation methods through oversubscription on multiple nodes.

The purpose of this work is to evaluate the efficacy of oversubscription, at the 1n, 2n, and 3n levels for n physical cores, on semi-direct MP2 methods within NWChem when using two and three Intel nodes. Semi-direct MP2 energy and gradient calculations were performed on chemical systems ranging from 824 to 1626 basis functions using the cc-pVDZ basis set. Wall times for semi-direct MP2 energies were reduced by as much as 36% using two nodes and 44% using three nodes compared to no oversubscription. Total energy consumed by the CPU and DRAM was also reduced by as much as 12% using two nodes and as much as 20% using three nodes when oversubscribing. MP2 gradient wall times improved by as much as 16% using two nodes and 18% using three nodes compared to execution at the 1n level; however, energy savings were insignificant. Intel performance-counter data show a strong correlation between total wall time saved and less time spent in the idle state, indicating a more efficient use of the processors when oversubscribing. © 2019 Wiley Periodicals, Inc.

Saving time and energy with oversubscription and semi-direct Møller-Plesset second order perturbation methods.

In this work, the effect of oversubscription is evaluated, via calling 2n, 3n, or 4n processes for n physical cores, on semi-direct MP2 energy and gradient calculations and RI-MP2 energy calculations with the cc-pVTZ basis using NWChem. Results indicate that on both Intel and AMD platforms, oversubscription reduces total time to solution on average for semi-direct MP2 energy calculations by 25-45% and reduces total energy consumed by the CPU and DRAM on average by 10-15% on the Intel platform. Semi-direct gradient time to solution is shortened on average by 8-15% and energy consumption is decreased by 5-10%. Linear regression analysis shows a strong correlation between time to solution and total energy consumed. Oversubscribing during RI-MP2 calculations results in performance degradations of 30-50% at the 4n level. © 2017 Wiley Periodicals, Inc.

Energy-Efficient Computational Chemistry: Comparison of x86 and ARM Systems.

The computational efficiency and energy-to-solution of several applications using the GAMESS quantum chemistry suite of codes is evaluated for 32-bit and 64-bit ARM-based computers, and compared to an x86 machine. The x86 system completes all benchmark computations more quickly than either ARM system and is the best choice to minimize time to solution. The ARM64 and ARM32 computational performances are similar to each other for Hartree-Fock and density functional theory energy calculations. However, for memory-intensive second-order perturbation theory energy and gradient computations the lower ARM32 read/write memory bandwidth results in computation times as much as 86% longer than on the ARM64 system. The ARM32 system is more energy efficient than the x86 and ARM64 CPUs for all benchmarked methods, while the ARM64 CPU is more energy efficient than the x86 CPU for some core counts and molecular sizes.

Design and Implementation of Scientific Software Components to Enable Multiscale Modeling: The Effective Fragment Potential (QM/EFP) Method.

The design and development of scientific software components to provide an interface to the effective fragment potential (EFP) methods are reported. Multiscale modeling of physical and chemical phenomena demands the merging of software packages developed by research groups in significantly different fields. Componentization offers an efficient way to realize new high performance scientific methods by combining the best models available in different software packages without a need for package readaptation after the initial componentization is complete. The EFP method is an efficient electronic structure theory based model potential that is suitable for predictive modeling of intermolecular interactions in large molecular systems, such as liquids, proteins, atmospheric aerosols, and nanoparticles, with an accuracy that is comparable to that of correlated ab initio methods. The developed components make the EFP functionality accessible for any scientific component-aware software package. The performance of the component is demonstrated on a protein interaction model, and its accuracy is compared with results obtained with coupled cluster methods.