DeepH-pack: a general-purpose neural network package for deep-learning electronic structure calculations
8 affiliations
- 1Department of Physics, Tsinghua University, Beijing, Beijing, China
- 2State Key Laboratory of Low Dimensional Quantum Physics, Tsinghua University, Beijing, Beijing, China
- 3Department of Materials Science and Engineering, Stanford University, Stanford, CA, USA
- 4Institute for Advanced Study, Tsinghua University, Beijing, Beijing, China
- 5Department of Physics, University of California, Berkeley, CA, USA
- 6Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA, USA
- 7Suzhou National Laboratory, Suzhou, Jiangsu, China
- 8Frontier Science Center for Quantum Information, Tsinghua University, Beijing, Beijing, China
Abstract
In computational physics and materials science, first-principles methods, particularly density functional theory, have become central tools for electronic structure prediction and materials design. Recently, rapid advances in artificial intelligence (AI) have begun to reshape the research landscape, giving rise to the emerging field of deep-learning electronic structure calculations. Despite numerous pioneering studies, the field remains in its early stages; existing software implementations are often fragmented, lacking unified frameworks and standardized interfaces required for broad community adoption. Here we present DeepH-pack, a comprehensive and unified software package that integrates first-principles calculations with deep learning. By incorporating fundamental physical principles into neural network design, such as the nearsightedness principle and the equivariance principle, DeepH-pack achieves robust cross-scale and cross-material generalizability. This allows models trained on small-scale structures to generalize to large-scale and previously unseen materials. The toolkit preserves first-principles accuracy while accelerating electronic structure calculations by several orders of magnitude, establishing an efficient and intelligent computational paradigm for large-scale materials simulation, high-throughput materials database construction, and AI-driven materials discovery.
Related
- VIBE-QC: an open-source, general-purpose electronic-structure code for molecules and solids developed with generative AI
- Biquaternion Algebra with Bilinear Multiplication: A General, Elegant, and Computationally Advantageous Framework for Relativistic Electronic Structure Calculations on CPUs and GPUs
- Reply to Comment on "Efficient implementation of the superposition of atomic potentials initial guess for electronic structure calculations in Gaussian basis sets"
- Nuclear moments, charge radii, and magnetization distribution parameters of Ag isotopes from laser spectroscopy and \textit{ab initio} electronic-structure calculations
- First-Principles Calculations of Structural, Electronic, Optical, Vibrational, Dielectric, and Elastic Properties of Tetragonal Mg2TiO4 and Mg2TiO4:Mn4+
BibTeX
@article{yang2026deephpack,
title = {{DeepH-pack: a general-purpose neural network package for deep-learning electronic structure calculations}},
author = {Li Yang and Yanzhen Wang and Boheng Zhao and Xiaoxun Gong and Yuxiang Wang and Zechen Tang and Zixu Wang and Zilong Yuan and Jialin Li and Minghui Sun and Zezhou Chen and Honggeng Tao and Baochun Wu and Yuhang Yu and He Li and Felipe H. da Jornada and Wenhui Duan and Yong Xu},
journal = {npj Computational Materials},
year = {2026},
month = jul,
doi = {10.1038/s41524-026-02219-2},
url = {https://doi.org/10.1038/s41524-026-02219-2},
note = {CC BY},
}
RIS
TY - JOUR
TI - DeepH-pack: a general-purpose neural network package for deep-learning electronic structure calculations
AU - Li Yang
AU - Yanzhen Wang
AU - Boheng Zhao
AU - Xiaoxun Gong
AU - Yuxiang Wang
AU - Zechen Tang
AU - Zixu Wang
AU - Zilong Yuan
AU - Jialin Li
AU - Minghui Sun
AU - Zezhou Chen
AU - Honggeng Tao
AU - Baochun Wu
AU - Yuhang Yu
AU - He Li
AU - Felipe H. da Jornada
AU - Wenhui Duan
AU - Yong Xu
PY - 2026
DA - 2026/07/09
JO - npj Computational Materials
DO - 10.1038/s41524-026-02219-2
UR - https://doi.org/10.1038/s41524-026-02219-2
AB - In computational physics and materials science, first-principles methods, particularly density functional theory, have become central tools for electronic structure prediction and materials design. Recently, rapid advances in artificial intelligence (AI) have begun to reshape the research landscape, giving rise to the emerging field of deep-learning electronic structure calculations. Despite numerous pioneering studies, the field remains in its early stages; existing software implementations are often fragmented, lacking unified frameworks and standardized interfaces required for broad community adoption. Here we present DeepH-pack, a comprehensive and unified software package that integrates first-principles calculations with deep learning. By incorporating fundamental physical principles into neural network design, such as the nearsightedness principle and the equivariance principle, DeepH-pack achieves robust cross-scale and cross-material generalizability. This allows models trained on small-scale structures to generalize to large-scale and previously unseen materials. The toolkit preserves first-principles accuracy while accelerating electronic structure calculations by several orders of magnitude, establishing an efficient and intelligent computational paradigm for large-scale materials simulation, high-throughput materials database construction, and AI-driven materials discovery.
SN - 2057-3960
ER -