ELSI: A unified software interface for Kohn–Sham electronic structure solvers

Yu, Victor; Corsetti, Fabiano; Garcı́a, Alberto; Huhn, William; Jacquelin, Mathias; Jia, Weile; Lange, Björn; Lin, Lin; Lu, Jianfeng; Mi, Wenhui; Seifitokaldani, Ali; Vázquez-Mayagoitia, Álvaro; Yang, Chao; Yang, Haizhao; Blüm, Volker

doi:10.1016/j.cpc.2017.09.007

Cited by 104 publications

(116 citation statements)

References 125 publications

Supporting

Mentioning

111

Contrasting

Order By: Relevance

“…[30,37,83] FHI-aims is a general-purpose electronic structure simulation code, offering proven scalability to thousands of atoms and on very large, conventional distributed-parallel high-performance computers. [84,77,85,86] The code achieves benchmark-quality accuracy for semi-local [87,17], hybrid [88,89,17], and many-body perturbative [88,90] levels of theory. The specific implementation described here helps unlock the potential of GPGPU architectures for a broad range of production simulations using semilocal DFT, including generalized gradient approximation (GGA) and meta-GGA exchangecorrelation functionals.…”

Section: Introductionmentioning

confidence: 99%

“…Shaded boxes indicate steps contributing to the actual computational workload. Yellow shading indicates steps that are subject to real-space GPU acceleration in this work, whereas gray shading indicates steps that are GPU-accelerated only partly, not at all, or that are handled by separate software components [92,77,79,78,86] outside the scope of this work. matrix…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

GPU acceleration of all-electron electronic structure theory using localized numeric atom-centered basis functions

Huhn

Lange

et al. 2020

Computer Physics Communications

Self Cite

View full text Add to dashboard Cite

We present an implementation of all-electron density-functional theory for massively parallel GPGPU-based platforms, using localized atom-centered basis functions and real-space integration grids. Special attention is paid to domain decomposition of the problem on non-uniform grids, which enables compute-and memory-parallel execution across thousands of nodes for realspace operations, e.g. the update of the electron density, the integration of the real-space Hamiltonian matrix, and calculation of Pulay forces. To assess the performance of our GPGPU implementation, we performed benchmarks on three different architectures using a 103-material test set. We find that operations which rely on dense serial linear algebra show dramatic speedups from GPGPU acceleration: in particular, SCF iterations including force and stress calculations exhibit speedups ranging from 4.5 to 6.6. For the architectures and problem types investigated here, this translates to an expected overall speedup between 3-4 for the entire calculation (including non-GPU accelerated parts), for problems featuring several tens to hundreds of atoms. Additional calculations for a 375-atom Bi 2 Se 3 bilayer show that the present GPGPU strategy scales for large-scale distributed-parallel simulations.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

GPU acceleration of all-electron electronic structure theory using localized numeric atom-centered basis functions

Huhn

Lange

et al. 2020

Computer Physics Communications

Self Cite

View full text Add to dashboard Cite

show abstract

“…Uncontrolled solver errors can be minimized on an equal footing with finite-basis and electron-correlation errors during the parameterization of a semiempirical model. Such a model would be associated with a specific solver algorithm, in contrast to the standard practice of numerical interoperability between solver algorithms as in the ELSI project [43].…”

Section: Discussionmentioning

confidence: 99%

Assessment of localized and randomized algorithms for electronic structure

Moussa

Baczewski

2019

Electron. Struct.

View full text Add to dashboard Cite

As electronic structure simulations continue to grow in size, the system-size scaling of computational costs increases in importance relative to cost prefactors. Presently, linear-scaling costs for three-dimensional systems are only attained by localized or randomized algorithms that have large cost prefactors in the difficult regime of low-temperature metals. Using large copper clusters in a minimal-basis semiempirical model as our reference system, we study the costs of these algorithms relative to a conventional cubic-scaling algorithm using matrix diagonalization and a recent quadratic-scaling algorithm using sparse matrix factorization and rational function approximation. The linear-scaling algorithms are competitive at the high temperatures relevant for warm dense matter, but their cost prefactors are prohibitive near ambient temperatures. To further reduce costs, we consider hybridized algorithms that combine localized and randomized algorithms. While simple hybridized algorithms do not improve performance, more sophisticated algorithms using recent concepts from structured linear algebra show promising initial performance results on a simple-cubic orthogonal tight-binding model.

show abstract

“…Calculations were performed with the Fritz Haber Institute ab initio molecular simulations (FHI-aims) package. [35][36][37] FHI-aims is based on numerical atomic orbitals and has been found to reach very high accuracies in ar ecent benchmark. [38] All structures in this study were optimized by using light orbitals and numeric integration settings until forces were lower than 0.001 eV À1 .F or periodic systems, a G-centred 6 6 1kpoints grid was used during geometry optimizations.…”

Section: Introductionmentioning

confidence: 99%

Clar Rules the Electronic Properties of 2D π‐Conjugated Frameworks: Mind the Gap

Strutyński

Mateo‐Alonso

Melle‐Franco

2020

Chemistry A European J

View full text Add to dashboard Cite

The key electronic properties of af amily of 2D frameworkss tructurally convergent with holey graphenes were studied. The bandgap of these materials decreases monotonically with size, showingacommon trend with anthracenes and kekulenes. This was rationalized by Clar's sextet rule, which reveals adirect relationship between the molecular systems and the 2D frameworks. In addition, ad etailed benchmark against experimental data showcasedt he high quality of the models, which reproduce accurately available electronic properties. Overall, it was shown that DFT can be used to screen and understand the intrinsic bandgapsa nd electrochemistry potentials for technological applications prior to the synthesis of p-conjugated porous materials.[c] Prof. A. Mateo-Alonso Ikerbasque, Basque Foundation for Science 48011B ilbao (Spain)Supporting information and the ORCID identification number(s) for the author(s) of this articlecan be found under: https://doi.

show abstract

ELSI: A unified software interface for Kohn–Sham electronic structure solvers

Cited by 104 publications

References 125 publications

GPU acceleration of all-electron electronic structure theory using localized numeric atom-centered basis functions

GPU acceleration of all-electron electronic structure theory using localized numeric atom-centered basis functions

Assessment of localized and randomized algorithms for electronic structure

Clar Rules the Electronic Properties of 2D π‐Conjugated Frameworks: Mind the Gap

Contact Info

Product

Resources

About