Showing 1–5 of 5 results for author: Ibrahim, K Z

Scalable Training of Trustworthy and Energy-Efficient Predictive Graph Foundation Models for Atomistic Materials Modeling: A Case Study with HydraGNN

Authors: Massimiliano Lupo Pasini, Jong Youl Choi, Kshitij Mehta, Pei Zhang, David Rogers, Jonghyun Bae, Khaled Z. Ibrahim, Ashwin M. Aji, Karl W. Schulz, Jorda Polo, Prasanna Balaprakash

Abstract: We present our work on developing and training scalable, trustworthy, and energy-efficient predictive graph foundation models (GFMs) using HydraGNN, a multi-headed graph convolutional neural network architecture. HydraGNN expands the boundaries of graph neural network (GNN) computations in both training scale and data diversity. It abstracts over message passing algorithms, allowing both reproduct… ▽ More We present our work on developing and training scalable, trustworthy, and energy-efficient predictive graph foundation models (GFMs) using HydraGNN, a multi-headed graph convolutional neural network architecture. HydraGNN expands the boundaries of graph neural network (GNN) computations in both training scale and data diversity. It abstracts over message passing algorithms, allowing both reproduction of and comparison across algorithmic innovations that define nearest-neighbor convolution in GNNs. This work discusses a series of optimizations that have allowed scaling up the GFMs training to tens of thousands of GPUs on datasets consisting of hundreds of millions of graphs. Our GFMs use multi-task learning (MTL) to simultaneously learn graph-level and node-level properties of atomistic structures, such as energy and atomic forces. Using over 154 million atomistic structures for training, we illustrate the performance of our approach along with the lessons learned on two state-of-the-art United States Department of Energy (US-DOE) supercomputers, namely the Perlmutter petascale system at the National Energy Research Scientific Computing Center and the Frontier exascale system at Oak Ridge Leadership Computing Facility. The HydraGNN architecture enables the GFM to achieve near-linear strong scaling performance using more than 2,000 GPUs on Perlmutter and 16,000 GPUs on Frontier. △ Less

Submitted 1 November, 2024; v1 submitted 12 June, 2024; originally announced June 2024.

Comments: 51 pages, 32 figures

MSC Class: 68T07; 68T09 ACM Class: C.2.4; I.2.11

Sparse-Stochastic Fragmented Exchange for Large-Scale Hybrid TDDFT Calculations

Authors: Mykola Sereda, Tucker Allen, Nadine C. Bradbury, Khaled Z. Ibrahim, Daniel Neuhauser

Abstract: We extend our recently developed sparse-stochastic fragmented exchange formalism for ground-state hybrid DFT (ngH-DFT) to calculate absorption spectra within linear-response time-dependent Generalized Kohn-Sham DFT (LR-GKS-TDDFT), for systems consisting of thousands of valence electrons within a grid-based/plane-wave representation. A mixed deterministic/fragmented-stochastic compression of the ex… ▽ More We extend our recently developed sparse-stochastic fragmented exchange formalism for ground-state hybrid DFT (ngH-DFT) to calculate absorption spectra within linear-response time-dependent Generalized Kohn-Sham DFT (LR-GKS-TDDFT), for systems consisting of thousands of valence electrons within a grid-based/plane-wave representation. A mixed deterministic/fragmented-stochastic compression of the exchange kernel, here using long-range explicit exchange functionals, provides an efficient method for accurate optical spectra. Both real-time propagation as well frequency-resolved Casida-equation-type approaches for spectra are presented, and the method is applied to large molecular dyes. △ Less

Submitted 25 February, 2024; originally announced February 2024.

Comments: 22 pages, 4 figures. First two authors are equal contributors to this work and designated as co-first authors

Velocity-gauge real-time time-dependent density functional tight-binding for large-scale condensed matter systems

Authors: Qiang Xu, Mauro Del Ben, Mahmut Sait Okyay, Min Choi, Khaled Z. Ibrahim, Bryan M. Wong

Abstract: We present a new velocity-gauge real-time, time-dependent density functional tight-binding (VG-rtTDDFTB) implementation in the open-source DFTB+ software package (https://dftbplus.org) for probing electronic excitations in large, condensed matter systems. Our VG-rtTDDFTB approach enables real-time electron dynamics simulations of large, periodic, condensed matter systems containing thousands of at… ▽ More We present a new velocity-gauge real-time, time-dependent density functional tight-binding (VG-rtTDDFTB) implementation in the open-source DFTB+ software package (https://dftbplus.org) for probing electronic excitations in large, condensed matter systems. Our VG-rtTDDFTB approach enables real-time electron dynamics simulations of large, periodic, condensed matter systems containing thousands of atoms with a favorable computational scaling as a function of system size. We provide computational details and benchmark calculations to demonstrate its accuracy and computational parallelizability on a variety of large material systems. As a representative example, we calculate laser-induced electron dynamics in a 512-atom amorphous silicon supercell to highlight the large periodic systems that can be examined with our implementation. Taken together, our VG-rtTDDFTB approach enables new electron dynamics simulations of complex systems that require large periodic supercells, such as crystal defects, complex surfaces, nanowires, and amorphous materials. △ Less

Submitted 21 May, 2024; v1 submitted 18 August, 2023; originally announced August 2023.

Comments: 23 pages, 6 figures

Journal ref: J. Chem. Theory Comput., 19, 22, 7989-7997 (2024)

Dynamic Mode Decomposition for Extrapolating Non-equilibrium Green's Functions Dynamics

Authors: Cian C. Reeves, Jia Yin, Yuanran Zhu, Khaled Z. Ibrahim, Chao Yang, Vojtech Vlcek

Abstract: The HF-GKBA offers an approximate numerical procedure for propagating the two-time non-equilibrium Green's function(NEGF). Here we compare the HF-GKBA to exact results for a variety of systems with long and short-range interactions, different two-body interaction strengths and various non-equilibrium preparations. We find excellent agreement between the HF-GKBA and exact time evolution in models w… ▽ More The HF-GKBA offers an approximate numerical procedure for propagating the two-time non-equilibrium Green's function(NEGF). Here we compare the HF-GKBA to exact results for a variety of systems with long and short-range interactions, different two-body interaction strengths and various non-equilibrium preparations. We find excellent agreement between the HF-GKBA and exact time evolution in models when more realistic long-range exponentially decaying interactions are considered. This agreement persists for long times and for intermediate to strong interaction strengths. In large systems, HF-GKBA becomes prohibitively expensive for long-time evolutions. For this reason, look at the use of dynamical mode decomposition(DMD) to reconstruct long-time NEGF trajectories from a sample of the initial trajectory. Using no more than 16\% of the total time evolution we reconstruct the total trajectory with high fidelity. Our results show the potential for DMD to be used in conjunction with HF-GKBA to calculate long time trajectories in large-scale systems. △ Less

Submitted 20 January, 2023; v1 submitted 10 November, 2022; originally announced November 2022.

An efficient basis set representation for calculating electrons in molecules

Authors: Jeremiah R. Jones, Francois-Henry Rouet, Keith V. Lawler, Eugene Vecharynski, Khaled Z. Ibrahim, Samuel Williams, Brant Abeln, Chao Yang, Daniel J. Haxton, C. William McCurdy, Xiaoye S. Li, Thomas N. Rescigno

Abstract: The method of McCurdy, Baertschy, and Rescigno, J. Phys. B, 37, R137 (2004) is generalized to obtain a straightforward, surprisingly accurate, and scalable numerical representation for calculating the electronic wave functions of molecules. It uses a basis set of product sinc functions arrayed on a Cartesian grid, and yields 1 kcal/mol precision for valence transition energies with a grid resoluti… ▽ More The method of McCurdy, Baertschy, and Rescigno, J. Phys. B, 37, R137 (2004) is generalized to obtain a straightforward, surprisingly accurate, and scalable numerical representation for calculating the electronic wave functions of molecules. It uses a basis set of product sinc functions arrayed on a Cartesian grid, and yields 1 kcal/mol precision for valence transition energies with a grid resolution of approximately 0.1 bohr. The Coulomb matrix elements are replaced with matrix elements obtained from the kinetic energy operator. A resolution-of-the-identity approximation renders the primitive one- and two-electron matrix elements diagonal; in other words, the Coulomb operator is local with respect to the grid indices. The calculation of contracted two-electron matrix elements among orbitals requires only O(N log(N)) multiplication operations, not O(N^4), where N is the number of basis functions; N = n^3 on cubic grids. The representation not only is numerically expedient, but also produces energies and properties superior to those calculated variationally. Absolute energies, absorption cross sections, transition energies, and ionization potentials are reported for one- (He^+, H_2^+ ), two- (H_2, He), ten- (CH_4) and 56-electron (C_8H_8) systems. △ Less

Submitted 22 January, 2016; v1 submitted 13 July, 2015; originally announced July 2015.

Comments: Submitted to JCP