Bulletin of the American Physical Society
APS March Meeting 2022
Volume 67, Number 3
Monday–Friday, March 14–18, 2022; Chicago
Session B48: Building the Bridge to Exascale: Applications and Opportunities in Materials, Chemistry, and Biology IFocus Recordings Available

Hide Abstracts 
Sponsoring Units: DCOMP DCP DPOLY DMP Chair: Jack Deslippe, LBNL Room: McCormick Place W471A 
Monday, March 14, 2022 11:30AM  12:06PM 
B48.00001: Towards a Realistic Description of H_{3}O+ and OH Transport through Confined Environments using Machine Learning and an OrderN Framework for CondensedPhase Hybrid Density Functional Theory Invited Speaker: Robert A Distasio By including a fraction of exact exchange (EXX), hybrid functionals reduce the selfinteraction error in semilocal density functional theory (DFT), and thereby furnish a more accurate and reliable description of the electronic structure in systems throughout chemistry, physics, and materials science. However, the high computational cost associated with hybrid DFT limits its applicability when treating largescale and complex condensedphase systems. To overcome this limitation, we have devised a highly accurate and linearscaling (orderN) approach based on a local representation of the occupied space that exploits sparsity when evaluating the EXX interaction in real space, and recently extended this framework to treat heterogeneous systems without the need for systemdependent parameters. In this work, we use this approach to generate highquality training data at the dispersioninclusive hybrid DFT level for training a reactive machinelearned potential to study how confinement affects the diffusion of H_{3}O+(aq) and OH(aq) at experimentally relevant length and time scales. To enable such largescale data generation, we have performed a comprehensive overhaul of our software to exploit nextgeneration highperformance computing architectures, including a threepronged strategy to improve the computation (including GPU acceleration), communication, and workload balance. With these developments, this work brings us closer to understanding H_{3}O+/OH transport through confined aqueous environments, which is of fundamental importance in the energy sciences (e.g., transport/conductivity in alkaline fuel cells). 
Monday, March 14, 2022 12:06PM  12:18PM 
B48.00002: Fueling a DataDriven Machine Learning Model for H_{3}O^{+} and OH^{} Transport through Confined Aqueous Environments: A HighThroughput OrderN Framework for CondensedPhase Hybrid Density Functional Theory at Work HsinYu Ko, Marcos F Calegari Andrade, Zachary M Sparrow, Brian G Ernst, Jalen A Harris, Robert A Distasio By including a fraction of exact exchange (EXX), hybrid functionals reduce the selfinteraction error in semilocal density functional theory (DFT), and thereby furnish a more accurate and reliable description of the electronic structure in systems throughout chemistry, physics, and materials science. In particular, it has been demonstrated that dispersioninclusive hybrid DFT can provide a semiquantitative description of H_{3}O^{+} and OH^{} transport in bulk aqueous solutions. However, the high computational cost associated with hybrid DFT limits its applicability when treating such largescale and complex condensedphase systems. To overcome this limitation, we have developed a highly accurate linearscaling (orderN) approach for treating finitegap (homogeneous and heterogeneous) systems without systemdependent parameters. Furthermore, we have implemented and devised a GPUaccelerated implementation of this framework to generate highquality dispersioninclusive hybrid DFT data for building a deep neural network potential for aqueous H_{3}O^{+} and OH^{} in bulk and confined environments. With these developments, this work brings us closer to understanding H_{3}O^{+} and OH^{} transport through confined aqueous environments, which is of fundamental importance in the energy sciences. 
Monday, March 14, 2022 12:18PM  12:30PM 
B48.00003: Optimization and performance of RMG DFTbased electronic structure software on exascale architectures Emil Briggs, Wenchang Lu, Jerry Bernholc Exascale computer architectures are close to deployment, with several systems expected to come online within the next 12 years. Exploiting their full computational power to study challenging scientific problems requires careful attention to machine design and its interaction with the algorithms used in scientific codes. The RMG software package for electronic structure calculations has been optimized for preexascale machines such as Summit at ORNL, and we describe the work being done to port RMG to exascaleclass machines. RMG uses a realspace formulation of the KohnSham equations that map well to distributed node architectures via domain decomposition. However, using mixed CPUGPU architectures efficiently requires careful work to hide or reduce the latencies associated with CPUGPU and internodedata transfers. We discuss the methods used to address these issues in base RMG and those emerging when implementing advanced features, such as hybrid functionals, semilocal pseudopotentials, and spinorbit coupling. Tests on AMD testbeds for the exascale Frontier show very promising performance. RMG source code and build scripts for preexascale Summit, Cray XEXK, clusters, Linux, and Windows workstations are at www.rmgdft.org, together with help files and examples. 
Monday, March 14, 2022 12:30PM  12:42PM 
B48.00004: CPUGPU optimization and performance of RMG linearscaling module with optimally localized orbitals Wenchang Lu, Emil Briggs, Jerry Bernholc The opensource RMG (RealSpace Multigrid) package solves the KohnSham equations directly on a realspace grid using multigrid acceleration. With a careful datastructure design, RMG is massively parallel on supercomputers with and without GPU accelerators. We recently released a new RMG module that can potentially lead to linear scaling with the system size, i.e., the number of atoms or electrons. In contrast to the main module, in which the wave functions are delocalized and directly represented on realspace grids, the localizedorbitals module expands the wave functions as a linear combination of strictlylocalized orbitals that are optimized variationally. For GPU acceleration, we implemented explicit memory management for multiple GPUs and CPU cores per node, which can be easily adapted to either CUDA (Nvidia) or HIP (AMD) programming environment. Timings on the new AMD testbed for the exascale Frontier show very good scalability and GPU speedup. 
Monday, March 14, 2022 12:42PM  12:54PM 
B48.00005: LargeScale Materials Science Codes Porting Strategies for Next Generation Exascale Architectures Mauro Del Ben, Steven G Louie, Jack R Deslippe Due to the intensive computational workload of their applications, materials science codes have been and still are among those which mostly benefit from leadership class HPC facilities. At the state of the art, graphics processing units (GPUs) dominate the HPC paradigm and force developers to actively maintain and optimize core compute kernels going forward. In this talk, we will focus on our experiences navigating this portability effort for the BerkeleyGW software package. BerkeleyGW is a massively parallel software package employed to study the excited state properties of electrons in materials by using the GW and the GW plus BetheSalpeter Equation (GWBSE) methods, and beyond. The code is capable of scaling out to tens of thousands of nodes and effectively utilizing strongscaling GPU architectures. We will discuss our experiences porting BerkeleyGW to three different GPU programming models (CUDA, OpenACC, OpenMP Target) and to various GPU vendor architectures, as well as challenges impeding true performance portability we encountered along the way. Special attention will be paid to code modernization practices which we found useful in the porting pipeline. 
Monday, March 14, 2022 12:54PM  1:06PM 
B48.00006: GPUAcceleration of LargeScale FullFrequency GW Calculations Victor Yu, Marco Govoni Manybody perturbation theory is a powerful method to simulate electronic excitations in molecules and materials starting from the output of density functional theory calculations. However, its widespread application to large systems has been hindered by the high computational cost. We present a GPU acceleration study of the fullfrequency GW method for periodic systems, as implemented in the WEST code [http://westcode.org]. We discuss the use of (1) optimized GPU libraries, e.g., cuFFT and cuBLAS, (2) a hierarchical parallelization strategy that minimizes CPUGPU, GPUGPU, and CPUCPU data transfer operations, (3) asynchronous GPU kernels that overlap with MPI communications, and (4) mixed precision in selected portions of the code. We demonstrate a substantial speedup of the GPUaccelerated version of WEST with respect to its CPU version, and we show good strong and weak scaling using up to 25,920 GPUs on the OLCF/Summit supercomputer. The GPU version of WEST yields electronic structures using the fullfrequency GW method for realistic nanostructures and interfaces comprising up to 10,368 electrons. This work was supported by MICCoM, as part of the Computational Materials Sciences Program funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences. 
Monday, March 14, 2022 1:06PM  1:18PM 
B48.00007: QMCPACK performance portability on NVIDIA and AMD GPUs Ye Luo, Peter Doak, Paul Kent As Exascale supercomputers are being deployed in U.S., QMCPACK (https://qmcpack.org) developers have migrated the code base to a performance portable implementation for science production on these powerful machines with outofbox experience. With a fresh design of code architecture, historically divergent code paths for CPUs and GPUs have been unified and a core set of features are available on all the computing platforms including CPUs and GPUs today. With portable OpenMP target offload programming model and high quality vendor linear algebra libraries, impressive performance has been achieved with minimal vendor specific customization needed. We show current performance for materials calculations on NVIDIA and AMD GPUs with a broad range of electron counts and analyze the remaining inefficiencies. 
Monday, March 14, 2022 1:18PM  1:30PM 
B48.00008: Using high accuracy manybody methods (QMC and sCI) to describe ground state and excited state properties of strongly correlated battery cathodes Anouar Benali, kevin E gasperich, Tomas Rojas, Vijay R Singh, Pallab Barai, Anh T Ngo, Hyeondeok Shin Since the commercialization of the first Lithium Ion Batteries (LIB) with LiCoO$_2$ as the active material in the cathodes, its electrochemical performance has been extensively studied both experimentally and theoretically. Increasing the capacity of LIB with high voltage leads to oxygen loss and surface reconstruction resulting in a rapid capacity fade and impedance growth. Using surface coating treatment stabilizes the surface and minimizes reactions with electrolytes but requires a good understanding of the electronic structure and properties of LiCoO$_2$. In the past decades, many studies using Density Functional Theory (DFT) corrected for strong correlation with an adhoc Hubbard U energy were published reproducing many important properties of the material. While DFT+U can reproduce a known property (band gap or lattice parameter), the value of U needs to be updated for each new property making the approach nonpredictive. In this talk we use a combination of Diffusion Monte Carlo (DMC) and select Configuration Interaction (sCI) for solids to describe the orbitals of LiCoO$_2$, leading to a better description of the band gaps of strongly correlated transition metal oxide materials and open the path to more reliable and trialwavefunction invariant DMC calculations. 
Monday, March 14, 2022 1:30PM  1:42PM 
B48.00009: Forces, stresses and related properties in solids by planewave auxiliaryfield quantum Monte Carlo Siyuan Chen, Fengjie Ma, Shiwei Zhang We present accurate calculations of interatomic force and stress in solid state systems, using the planewave auxiliaryfield quantum Monte Carlo (PWAFQMC) method [1] . AFQMC has been shown to be an excellent manybody total energy method. Computation of observables other than the groundstate energy requires backpropagation [2], which we have implemented in the PWAFQMC framework and to compute accurate charge densities [3]. Here we present results on computing derivatives of the total energy, including forces and stresses. Accurate AFQMC interatomic forces and stresses can be applied for a full geometry optimization in solids, which we demonstrate in the silicon betatin structure and molybdenum disulfide (MoS_{2}) monolayer. Further, we generalize the correlatedsampling technique [4] to compute observables, and demonstrate it by computing the phonon spectrum from the force fields at a low cost. This paves the way for ab initio manybody computation of thermodynamic properties. 
Monday, March 14, 2022 1:42PM  1:54PM 
B48.00010: Ab initio Calculations in Atoms, Molecules, and Solids, Treating SpinOrbit Coupling and Electron Interaction on Equal Footing Brandon Eskridge, Henry Krakauer, Hao Shi, Shiwei Zhang Understanding the interplay between electronelectron interaction and spinorbit coupling in molecules and materials is key to answering various fundamental and technological questions. An unbiased theoretical treatment requires that material specificity, electron correlation, and spinorbit coupling (SOC) be captured accurately and on equal footing. We have incorporated explicit, nonperturbative treatment of spinorbit coupling into ab initio auxiliaryfield quantum Monte Carlo (AFQMC) calculations. The approach allows a general computational framework for molecular and bulk systems in which materials specificity, electron correlation, and spinorbit coupling effects can be captured accurately, with favorable computational scaling versus system size. We adopt relativistic effectivecore potentials which have been obtained by fitting to fully relativistic data and which have demonstrated a high degree of reliability and transferability in molecular systems. This results in a 2component spincoupled Hamiltonian, which is then treated by generalizing the ab initio AFQMC approach. We demonstrate the method by computing the electron affinity in Pb, the bond dissociation energy in Br_{2} and I_{2}, and solid Bi. 
Monday, March 14, 2022 1:54PM  2:06PM 
B48.00011: Electronic structure calculations at finite temperature using the piecewise interaction picture density matrix quantum Monte Carlo approach William Z Van Benschoten, James J Shepherd In this work, we present a modification to the propagator in density matrix quantum Monte Carlo methods (DMQMC) which is a method to sample the Nbody density matrix for an electronic Hamiltonian. Starting from the interaction picture (IP) requires IPDMQMC to sample only one temperature during a single calculation; this was developed to alleviate the undersampling of the important energy states. The new approach combines the IPDMQMC and DMQMC propagators in a piecewise fashion (PIPDMQMC) allowing for nearcontinuous temperature sampling. We benchmark this method by comparing to the sum over full configuration interaction (FCI) states, IPDMQMC, and DMQMC. We find equivalent or improved energy estimates for the benchmark systems. We then use initiator PIPDMQMC to simulate the water and methane molecules in the ccpVDZ basis. Finally, we compare the cost of this method to that of the original IPDMQMC method. We believe this method will extend the size of systems which can be accurately treated with finite temperature and will provide useful comparison data for other finite temperature methods. 
Monday, March 14, 2022 2:06PM  2:18PM 
B48.00012: Rapid Generation of Optimal Generalized MonkhorstPack Grids Yunzhe Wang, Pandu Wisesa, Adarsh Balasubramanian, Shyam Dwaraknath, Tim Mueller The calculation of properties of crystalline materials can often be accelerated by using kpoint grids to estimate an integral in reciprocal space. We have shown that a generalization of the widelyused MonkhorstPack method for kpoint grid generation can reduce the number of symmetrically distinct kpoints required to reach a given level of accuracy, and thus accelerate computational throughput, by roughly a factor of two for wellconverged density functional theory calculations. To facilitate the widespread adoption of this approach, we have developed algorithms that enable dynamic generation of optimal generalized MonkhorstPack grids within seconds. We present these algorithms and several software tools in which they are implemented, including a C++ library with a Python interface designed for integration with thirdparty software packages. 
Follow Us 
Engage
Become an APS Member 
My APS
Renew Membership 
Information for 
About APSThe American Physical Society (APS) is a nonprofit membership organization working to advance the knowledge of physics. 
© 2024 American Physical Society
 All rights reserved  Terms of Use
 Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 207403844
(301) 2093200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 5914000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 200452001
(202) 6628700