Bulletin of the American Physical Society
APS April Meeting 2022
Volume 67, Number 6
Saturday–Tuesday, April 9–12, 2022; New York
Session S07: Simulations Methods and ImplementationRecordings Available
|
Hide Abstracts |
Sponsoring Units: DCOMP Chair: Charles Gammie, University of Illinois at Urbana-Champaign Room: Salon 4 |
Monday, April 11, 2022 1:30PM - 1:42PM |
S07.00001: Towards Quantum Compression in High-Energy Physics Data Santiago R Giner Olavarrieta, Jeffrey P Lazar, Giancarlo Gatti High-energy physics experiments produce a vast amount of data. For instance, the Large Hadron Collider alone generates about 250 TB of data every day. This raises the obvious challenge of efficiently storing and operating on this information. This task will only become more difficult as we move towards the next generation of experiments and observatories, which will produce an order of magnitude more data. Even though quantum computing is in its nascency, it has the potential to revolutionize information technology and therefore address many of the aforementioned challenges. Specifically, the development of persistent quantum random access memory would allow us to store data in the exponentially large Hilbert space. In this contribution, I will present a novel high-density encoding of high-energy physics data as well as prospects for operating on such densely encoded data. |
Monday, April 11, 2022 1:42PM - 1:54PM |
S07.00002: Electronic structure of semiconductor nanoparticles from stochastic evaluation of imaginary-time path integral Andrei B Kryjevski The fermion sign problem, when severe, prevents the computation of physical quantities of a system of interacting |
Monday, April 11, 2022 1:54PM - 2:06PM |
S07.00003: Variational quantum algorithms for PDE measurement extraction Pete B Rigas Classical-quantum hybrid algorithms have garnered attention for their ability to combine classical simulations of minimization with quantum simulation of cost functions, which are broadly characterized by various amalgamations of quantum and classical computing protocols for obtaining readout of solutions. In particular, one recent VQA due to Lubasch et al in 2019 exhibits a theoretical and empirically focused study for obtaining solutions to the Schrodinger and Inviscid Burgers equations, through optimizing for the ground state of cost functions parametrized in each PDE that is being solved. With approximations of nonlinearities that are encoded with expectation values of the cost function, the authors report of high fidelity as their computational quantum-driven methods are capable of reliably predicting classically determined solutions. To determine whether the VQA is suitable for more complicated nonlinearities of other PDEs, in addition to corresponding solutions that are posed with different initial conditions, we discuss investigations of other cost functions that can be clasically minimized. For each nonlinear problem of interest, the quantum state encoded by the cost function is representative of new physical phenomena that can be studied from time-evolving quantum states that are represented with sequences of unitary gates acting on qubits initialized within quantum circuit registers. To avoid encountering barren plateaus and inescapable minima in the optimization landscape that have been throughouly characterized when time-evolving quantum representations of solutions by other areas of the quantum computing community, we form mathematically precise characterizations of the performance of the algorithm for other initial value problems, through simulation of several nonlinear behaviors. To study other nonlinearities, we formulate cost functions dependent upon time-evolved variational states of solutions, and form comparisons with clasically determined solutions. |
Monday, April 11, 2022 2:06PM - 2:18PM |
S07.00004: Efficient Parallel Numerical Simulations of the Einstein Equations in Spherical Coordinates Liwei Ji We present a new parallel filter that we use to ameliorate the severe |
Monday, April 11, 2022 2:18PM - 2:30PM |
S07.00005: The Impact of New and Emerging Architectures on Computational Physics Eva Siegmann We present experiences with Ookami, a testbed machine providing open access to the A64FX processor developed by Fujitsu for the Japanese path to exascale computing. We provide an overview of the project with details of the user base and the diverse applications of users. We also present a case study of experience with running the community simulation code FLASH, developed at the University of Chicago for multi-scale multi-physics applications. We present results with different compilers and other software, our initial scaling results, and attempts to utilize the A64FX's SVE instructions and NUMA architecture. We present results from several applications including scaling studies, compiler comparisons, and tuning for SVE and communication patterns. We also share best practices for getting the most out of this new architecture. |
Monday, April 11, 2022 2:30PM - 2:42PM |
S07.00006: Educating HPC Users in the Use of Advanced Computing Technology Alan C Calder, Eva Siegmann, Catherine Feldman, Robert Harrison We present a multi-modal approach to educating users of Ookami, an HPE Apollo 80 testbed featuring the Fujitsu A64FX processor that is open to the community for porting and testing applications. Achieving good performance on this Arm-based technology necessitates familiarity with details of computer architecture, performance analysis and modeling, and high-performance programming models as well as basic concepts such as vectorization, pipelining, latency/bandwidth, threads, and non-uniform memory access, topics often omitted in introductory programming courses. We employ multiple approaches to education that emphasize (online) personal interactions and transfer of skills for our global user community. Documentation is organized around best-practices and FAQs; twice-weekly "office hours" via Zoom enable deep dives by both the team and the user community; a Slack channel provides both real time and archived answers and discussions; and we host workshops and webinars. Using these tools for education rather than just for project communication makes them more effective and contributes to community success. |
Monday, April 11, 2022 2:42PM - 2:54PM |
S07.00007: Relativistic blast wave and single neutron star tests with the Nmesh Code Ananya Adhikari, Wolfgang H Tichy, Liwei Ji, Alireza Rashti We present some recent tests that we have performed with the Nmesh code. These include the propagation of relativistic blast waves and evolution of a single neutron star. We propagate the relativistic blast waves entirely with a discontinuous Galerkin (DG) method, and also entirely with a finite volume (FV) method. For the case of evolving the blast wave with the DG method, we apply the MRS limiter to the primitive variables of relativistic hydrodynamics and compare with the results when we apply the limiter to the conservative variables. For the FV method evolution of the blast wave, we apply a WENO scheme implementation that we have developed and we explore the results. For the neutron star, we use the DG method in the interior of the star where the density is smooth, and the FV method near the star surface. |
Monday, April 11, 2022 2:54PM - 3:06PM |
S07.00008: Simulation of polaron transport in defected spinel oxides Maytal Toroker The small-polaron hopping model has been used for several decades for modeling electronic charge transport in oxides. Despite its significance, the model was developed for binary oxides, and its accuracy has not been rigorously tested for higher-order oxides. To investigate this issue, we chose the MnxFe3-xO4 spinel system, which has exciting electrochemical and catalytic properties, and mixed cation oxidation states that enable us to examine the mechanisms of small-polaron transport. Using a combination of experimental results and DFT+U calculations, we find that the charge transport occurs only between like-cations (Fe/Fe or Mn/Mn). And due to asymmetric hopping barriers and formation energies, we find that the polaron is energetically preferred to the polaron, resulting in an asymmetric contribution of the Mn/Mn pathways. |
Monday, April 11, 2022 3:06PM - 3:18PM |
S07.00009: Parallel I/O optimization for micrometer-scale atomistic calculations under extreme conditions Sanghyuk Yoo, Seunghwan Oh, Younggak Shin, Ohkyoung Kwon, Keonwook Kang, Byeongchan Lee Atomistic calculations is an essential means for materials under extreme conditions; for example, fusion reactors are required to be designed without the necessary environment to test them. To better complement experiments and make predictions more useful, however, improvements are necessary in a quantitative sense against experimental measurements as well as theoretical references such as first principles calculations. One way is to develop accurate interatomic potentials, and the other is to break scaling limits. In this study, we focus on how to extend molecular dynamics beyond the conventional length scale to what we call a micron virtual specimen that contains tens of billions of atoms. Common problems such as a stall due to I/O bandwidth arise for such large calculations necessarily running on 68K+ CPU cores, which can be avoided by assigning only a part of the allocated processes to I/O while the rest to calculations. We will report the current development status along with the preliminary results for irradiated-damage simulations on KISTI Nurion. |
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