Related Experiment Video
Updated: Nov 23, 2025

Computation of Atmospheric Concentrations of Molecular Clusters from ab initio Thermochemistry
Published on: April 8, 2020
Accurate Reduced-Cost CCSD(T) Energies: Parallel Implementation, Benchmarks, and Large-Scale Applications.
László Gyevi-Nagy1, Mihály Kállay1, Péter R Nagy1
1Department of Physical Chemistry and Materials Science, Budapest University of Technology and Economics, P.O. Box 91, H-1521 Budapest, Hungary.
We combined cost-reducing Frozen Natural Orbital (FNO) and Natural Auxiliary Function (NAF) methods with Coupled-Cluster Singles, Doubles, and Perturbative Triples [CCSD(T)] calculations. This approach achieves high accuracy for large molecular systems, significantly reducing computational cost.
Area of Science:
- Computational Chemistry
- Quantum Chemistry
- Method Development
Background:
- Coupled-Cluster Singles, Doubles, and Perturbative Triples [CCSD(T)] is a high-accuracy quantum chemistry method.
- Computational cost limits the application of canonical CCSD(T) to small molecular systems.
- Frozen Natural Orbital (FNO) and Natural Auxiliary Function (NAF) approaches offer significant cost reductions.
Purpose of the Study:
- To implement and benchmark combined Frozen Natural Orbital (FNO) and Natural Auxiliary Function (NAF) approximations with Coupled-Cluster Singles, Doubles, and Perturbative Triples [CCSD(T)].
- To assess the accuracy and scalability of the developed FNO-CCSD(T) method for various chemical properties and system sizes.
- To extend the applicability of accurate electronic structure calculations to larger and more complex molecular systems.
Main Methods:
- Integration of FNO and NAF approximations into existing CCSD(T) codes.
- Implementation of OpenMP parallelism, integral-direct density-fitting, and checkpointing for efficiency.
- Benchmarking against canonical CCSD(T) for reaction, atomization, and ionization energies of closed- and open-shell species.
Main Results:
- FNO-CCSD(T) maintains chemical accuracy (within 1 kJ/mol) compared to canonical CCSD(T) for systems up to 43 atoms.
- Computational cost is reduced by up to an order of magnitude, enabling calculations for systems with 50-75 atoms.
- The method demonstrates excellent parallel performance and high peak performance utilization on hundreds of cores.
Conclusions:
- The developed FNO-CCSD(T) method provides a computationally feasible route to "gold standard" accuracy for significantly larger systems.
- This advancement expands the accessible chemical space for high-accuracy electronic structure calculations.
- Potential applications include benchmarking other computational methods and studying complex chemical reactions and interactions.
More Related Videos
13:56Probe Type II Band Alignment in One-Dimensional Van Der Waals Heterostructures Using First-Principles Calculations
Published on: October 12, 2019
05:37Rapid in-silico Battery Electrolyte Electrochemical Reaction Generation using 3T-VASP Multi-Scale Energy Minimization
Published on: August 22, 2025
Related Concept Videos
Parallel Processing
Fast Decoupled and DC Powerflow