KOR

e-Article

Materials and Molecular Modeling at the Exascale
Document Type
Periodical
Source
Computing in Science & Engineering Comput. Sci. Eng. Computing in Science & Engineering. 24(1):36-45 Jan, 2022
Subject
Computing and Processing
Bioengineering
Communication, Networking and Broadcast Technologies
Computational modeling
Software
Discrete Fourier transforms
Exascale computing
Dynamics
Quantum mechanics
Quantum computing
Language
ISSN
1521-9615
1558-366X
Abstract
The transition to exascale computing will make possible simulations of unprecedented accuracy and complexity. We focus on materials and molecular modeling (MMM) aspiring to high fidelity, in silico experiments on complex systems of technological interest. This progress will present unprecedented challenges to the software used, especially how to exploit the huge degree of parallelism and the associated problems of creating effective workflows and data management on such platforms. Within the U.K.’s ExCALIBUR computing initiative, our U.K.-led MMM Design and Development Working Group has worked with the broad MMM community to identify high-priority applications that will drive future exascale software developments. We present an overview of selected case studies that pose new methodological challenges on exascale platforms and discuss the requirements, software challenges, and impact of each application area.