Scalability of Quantum Chemical Calculations of Crystalline Materials Using "Tornado" Supercomputer in South Ural State University
DOI:
https://doi.org/10.14529/cmse220304Keywords:
digital twin, quantum chemical calculations, functional materials, scalabilityAbstract
Technical results of multiscale modeling of crystalline materials of different structure and composition has been presented and discussed. The scalability of different types of quantum chemical calculations using resources of "Tornado" supercomputer has been studied. Such tasks are extremely relevant in the field of modeling of structure and properties of crystals and carbon materials. The range of scalability has been reported for the systems of different size, composition, symmetry and level of modeling in the framework of density functional theory with atomic basis sets. A list of recommendations has been formulated presenting the optimal parameters for different types of material science tasks on the way to digital twin design. Analysis of calculation time for different systems and calculation types has been performed. A list of simple systems such as 3D silicon-substituted graphite as well as complex systems consisting of carbon surfaces (nanotube and graphene layer) and Bisphenol A diglycidyl ether as molecule on a surface. It is revealed that for large systems the most critical condition is the increase of the size of temporary files with the increase of number of nodes leading to possible failure of the calculation. It was shown that the scalability of the calculations of vibration properties of crystals is significantly lower than of the search of energetically preferable structure no matter how many atoms are located in the elementary part of the cell of computed structure.References
Xiang F., Zhang Z., Zuo Y., Tao F. Digital Twin Driven Green Material Optimal-Selection towards Sustainable Manufacturing. Procedia CIRP. 2019. Vol. 81, no. 1. P. 1290–1294. DOI: 10.1016/j.procir.2019.04.015.
Noel Y., D’Arco P., Demichelis R., et al. On the use of symmetry in the ab initio quantum mechanical simulation of nanotubes and related materials. Journal of Computational Chemistry. 2010. Vol. 31, no. 4. P. 855–862. DOI: 10.1002/jcc.21370.
Civalleri B., D’Arco P., Orlando R., et al. Hartree-Fock geometry optimisation of periodic systems with the CRYSTAL code. Chemical Physics Letters. 2001. Vol. 348, no. 1. P. 131–138. DOI: 10.1016/s0009-2614(01)01081-8.
Dovesi R., Pascale F., Civalleri B., et al. The CRYSTAL code, 1976–2020 and beyond, a long story. Journal of Chemical Physics. 2020. Vol. 152, no. 1. P. 204111. DOI: 10.1063/5.0004892.
Bandura A., Evarestov R. Nonempirical methods of modeling of crystals with atomic basis set using parallel calculations. Saint Petersburg: Publishing of Saint Petersburg University, 2004. 228 p. (in Russian).
Pascale F., Zicovich-Wilson C., Gejo F.L., et al. The calculation of vibrational frequencies of crystalline compounds and its implementation in the CRYSTAL code. Chemical Physics Letters. 2004. Vol. 25, no. 6. P. 888–897. DOI: 10.1002/jcc.20019.
Lohse M.S., Bein T. Covalent Organic Frameworks: Structures, Synthesis, and Applications. Advanced Functional Materials. 2018. Vol. 28, no. 33. P. 1705553. DOI: 10.1002/adfm.201705553.
Dolganina N., Ivanova E., Bilenko R., Rekachinsky A. HPC Resources of South Ural State University. Parallel Computational Technologies. PCT 2022. Communications in Computer and Information Science. Springer, 2022. DOI: 10.1007/978-3-031-11623-0_4.
