An exciting extravagant extraordinarily exceptional exascale experience - from water and bonding, to 27,000 GPUs and beyond!
Date
2022-12
Authors
Galvez Vallejo, Jorge Luis
Major Professor
Advisor
Gordon, Mark
Huang, Wenyu
Evans, James
Song, Xueyu
Windus, Theresa
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The work described in this dissertation covers the development of electronic structure software geared towards high performance computing applications, specifically aimed at the use of Graphics Processing Units (GPUs).
The first chapter serves as a general introduction into the field of quantum chemistry, followed by a brief coverage of high performance computing emphasizing the challenges and obstacles of writing quantum chemistry programs geared towards high performance. The second through sixth chapters focus on the development of GPU accelerated programs such as two electron integrals, the Self-Consistent-Field (SCF) algorithm, and their further scaling up to the Summit supercomputer at Oak Ridge National Laboratory. The seventh chapter covers the development of a novel task based approach to the Restricted-Hartree-Fock (RHF) method in the novel Julia programming language. Chapters eight through ten cover the applications of computational chemistry in the fields of chemical bonding, fuel aging, and heterogeneous catalysis. Chapter eight discusses the bonding motifs in different sized small water clusters, with a deep emphasis into the hydrogen bonds that keep the system together. Chapter nine, dwells into the possible degradation pathways of solid propellants based on Hydroxy-terminated-polybutadiene (HTPB) and their potentialk thermochemical properties. Chapter ten describes the use of heteropolyacids, specifically phosphotungstic acid (HPW) as a heterogenous catalyst for endothermic fuel applications. Chapter eleven offers general conclusions.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
dissertation