Bulletin of the American Physical Society
APS March Meeting 2023
Volume 68, Number 3
Las Vegas, Nevada (March 5-10)
Virtual (March 20-22); Time Zone: Pacific Time
Session F03: Electronic Structure in Open Science IFocus
|
Hide Abstracts |
Sponsoring Units: DCP Chair: Theresa Windus, Ames Laboratory Room: Room 126 |
Tuesday, March 7, 2023 8:00AM - 8:36AM |
F03.00001: Standardizing Code Interoperability Through the MolSSI Driver Interface Project Invited Speaker: Taylor A Barnes The MolSSI Driver Interface (MDI) Project provides a standardized API for fast, on-the-fly communication between software in the computational molecular sciences domain. The API is sufficiently general to support a wide variety of techniques, including QM/MM, ab initio MD, machine learning, advanced sampling, and path integral MD, while also being straightforwardly extensible. Furthermore, the MDI project's flexible API enables codes to interact at multiple different points throughout a simulation; this effectively allows MDI-enabled codes to insert callback functions that can modify behavior at a lower level than what is possible through a typical library API. The details of inter-code communication are handled by the MDI Library, which supports tight coupling via MPI, TCP/IP sockets, or through a plugin interface. With support in a growing number of codes, MDI offers developers an opportunity to create general interfaces that ensure their codes can be easily interoperated with other software in modular manner. |
Tuesday, March 7, 2023 8:36AM - 8:48AM |
F03.00002: SHRED: An open-source DFT code for exascale and matter in extreme conditions Alexander J White Real space and planewave based Kohn-Sham Density Functional Theory codes are critical tools for studying condensed matter, chemical, material, and plasma physics. However, a large basis and the need to orthogonalize large numbers of orbitals/bands leads to computational complexity that scales cubically in both system size and temperatures, in the electron-volt regime. Additionally, significant communication bottlenecks limit parallel scaling across many nodes and/or GPU’s. In this talk, we present the SHRED (Stochastic and Hybrid Representation for Electronic structure by Density functional theory) code which utilizes alternative linear-scaling stochastic, mixed stochastic-deterministic, and orbital-free DFT and TD-DFT algorithms to circumvent orbital orthogonalization and achieve significant acceleration of calculations in a range of simulations. Newly implemented PAW pseudopotentials (based on Abinit’s LibPAW library), progress in GPU acceleration and new correlated sampling techniques, and applications to warm dense matter will be highlighted. |
Tuesday, March 7, 2023 8:48AM - 9:24AM |
F03.00003: Composing and decomposing quantum chemistry software: Adventures with Psi4 and QCArchive Invited Speaker: Lori A Burns The Psi4 quantum chemistry (QC) program maintains a full-featured core but also has accumulated interfaces with over a dozen specialist community libraries. Recently, Psi4 has added to its ecosystem by divesting some of its post-processing and interface features into independent projects knitted together by QCSchema, a collection of JSON-based communication structures governed by Pydantic, and run by QCEngine, an I/O standardizer for two dozen chemistry projects. QCSchema and QCEngine are part of the Quantum Chemistry Archive (QCA), an umbrella project by the Molecular Sciences Software Institute (MolSSI) to accumulate and serve high-volume quantum chemical results for the community. Building upon QCA, Psi4's newly reworked outer Python layer facilitates high-throughput computing of naturally parallel procedures such as composite methods or many-body routines with minimal changes to the input. Capabilities to call Psi4 and other QC programs through increasingly uniform input suitable for software generation will also be presented. |
Tuesday, March 7, 2023 9:24AM - 9:36AM |
F03.00004: A Multimer Embedding Approach for Energies, Structures, and Vibrational Properties of Molecular Crystals Johannes Hoja, Alexander List, A. Daniel Boese Accurate calculations of molecular crystals are crucial for crystal engineering and drug design. However, high-level electronic structure calculations are often prohibitively expensive for relevant systems. Such expensive periodic calculations can be circumvented by the usage of embedding methods. For instance, the periodic result of a high-level method can be approximated by a subtractive embedding scheme, in which a fully periodic calculation is only performed using a less expensive lower-level method and then monomer energies and dimer interaction energies are replaced by those of the high-level method. Herein, we present such a multimer embedding approach containing up to trimer interactions for energies, structures, and vibrational properties of molecular crystals. We demonstrate the high accuracy of this approach for the X23 benchmark set of molecular crystals by approximating a periodic hybrid density functional (PBE0+MBD) by embedding multimers into less expensive calculations using a generalized-gradient approximation (GGA) functional (PBE+MBD). We show that trimer interactions are crucial for accurate lattice energies and cell volumes, while harmonic vibrational properties can already be well captured at the dimer level. Finally, we use this multimer embedding approach to also incorporate anharmonic effects. |
Tuesday, March 7, 2023 9:36AM - 10:12AM |
F03.00005: Regularized CASPT2: an intruder-state-free approach Invited Speaker: Roland Lindh In this work we present a new approach to fix the intruder state problem (ISP) in CASPT2 based on σp regularization. The resulting σp-CASPT2 method is compared to previous techniques, namely the real and imaginary level shifts, on a theoretical basis and by performing a series of systematic calculations. The analysis is focused on two aspects, the effectiveness of σp-CASPT2 in removing the ISP and the sensitivity of the approach with respect to the input parameter. We found that σp-CASPT2 compares favorably with respect to previous approaches, and that different versions, σ1-CASPT2 and σ2-CASPT2, have different potential application domains. This analysis also reveals the unsuitability of the real level shift technique as a general way to avoid the intruder state problem. |
Tuesday, March 7, 2023 10:12AM - 10:24AM |
F03.00006: Optimizing quantum measurements of electronic Hamiltonians in the variational quantum eigensolver Seonghoon Choi, Tzu-Ching Yen, Ignacio Loaiza, Artur F Izmaylov Efficient measurement of the expectation value of the molecular electronic Hamiltonian is crucial for the variational quantum eigensolver to be practical. A widely used strategy is to partition the Hamiltonian into linear combinations of operators that can be measured simultaneously (fragments). The total number of measurements required to obtain the Hamiltonian expectation value is then proportional to the sum of fragment variances. Here, we introduce two new methods for lowering the fragments’ variances by exploiting the flexibility in the fragments’ form. One approach is based on adding Pauli products (ghosts) that are compatible with members of multiple fragments such that all ghosts sum to zero [1], while the other approach uses the idempotency of the fermion occupation number operators to turn some parts of two-electron fragments into one-electron fragments, which are then partially collected into a purely one-electron fragment [2]. In both approaches, the modifications do not affect the total Hamiltonian expectation value but have non-vanishing contributions to the variance of each fragment. The proposed algorithms minimize individual fragment variances using a classically efficient approximation of the quantum wavefunction for variance estimations. Numerical tests on several molecules show that the algorithms can lower the number of measurements by more than an order of magnitude. |
Tuesday, March 7, 2023 10:24AM - 11:00AM |
F03.00007: Effective Fragment Molecular Orbital (EFMO) Method: Ab Initio-Based Fragmentation Approach for Large Systems Invited Speaker: Peng Xu Accurately and efficiently modeling nanosized molecular clusters and beyond can now be achieved with an ab initio-based fragmentation approach, the effective fragment molecular orbital (EFMO) method. The computational bottlenecks of EFMO calculations are identified and the codes are ported to GPUs. The accuracy and efficiency of the method are substantially improved with several advancements including a multi-layer implementation. This talk will highlight several progresses made in order to simulate the reaction pathways occurring in mesoporous silica nanoparticles in the solvent. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit 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 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700