SEARCH · Search NASA

Results for “Scaling behavior”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 181 records · Page 10

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Scaling neural simulations in STACS

Abstract As modern neuroscience tools acquire more details about the brain, the need to move towards biological-scale neural simulations continues to grow. However, effective simulations at scale remain a challenge. Beyond just the tooling required to enable parallel execution, there is also the unique structure of the synaptic interconnectivity, which is globally sparse but has relatively high connection density and non-local interactions per neuron. There are also various practicalities to consider in high performance computing applications, such as the need for serializing neural networks to support potentially long-running simulations that require checkpoint-restart. Although acceleration on neuromorphic hardware is also a possibility, development in this space can be difficult as hardware support tends to vary between platforms and software support for larger scale models also tends to be limited. In this paper, we focus our attention on Simulation Tool for Asynchronous Cortical Streams (STACS), a spiking neural network simulator that leverages the Charm++ parallel programming framework, with the goal of supporting biological-scale simulations as well as interoperability between platforms. Central to these goals is the implementation of scalable data structures suitable for efficiently distributing a network across parallel partitions. Here, we discuss a straightforward extension of a parallel data format with a history of use in graph partitioners, which also serves as a portable intermediate representation for different neuromorphic backends. We perform scaling studies on the Summit supercomputer, examining the capabilities of STACS in terms of network build and storage, partitioning, and execution. We highlight how a suitably partitioned, spatially dependent synaptic structure introduces a communication workload well-suited to the multicast communication supported by Charm++. We evaluate the strong and weak scaling behavior for networks on the order of millions of neurons and billions of synapses, and show that STACS achieves competitive levels of parallel efficiency.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2025

Rosenbluth-like separation of the $J/ψ$ near-threshold photoproduction: An access to the gluon gravitational form factors at high t

Here, we perform analysis of the near-threshold $J/\psi $ photoproduction data off the proton based on two theoretical approaches, GPD \cite{Guo3} and holographic \cite{Zahed2}, that represent the differential cross sections as powers of the skewness parameter with coefficients that depend only on the momentum transfer $t$. This allows to separate kinematically the corresponding coefficient functions, in much the same way as this is done for the electric and magnetic form factors using the Rosenbluth separation. We examine the independence of the extracted functions with the photon beam energy. These functions, under additional assumptions, are related to the proton's gluon Gravitational Form Factors (gGFFs). We compare the extracted functions with lattice calculations of the gGFFs in the region of $0.5<|t|<2$~GeV$^{2}$, where they overlap. Such analysis demonstrates the possibility of extracting some combinations of the gGFFs from the data at high $t$, complementary to the lattice calculations available in the low $t$ region. However, higher statistics are needed to more accurately check the predicted scaling behavior of the data and compare with the lattice results, thus testing and comparing the theoretical assumptions used in the GPD and holographic models.

Pentchev, Lubomir [Thomas Jefferson National Accel↗

DOE OSTI2026

Efficient many-jet event generation with flow matching

We apply for the first time, to the best of our knowledge, the flow matching method to the problem of phase-space sampling for event generation in high-energy collider physics. By training the model to remap the random numbers used to generate the momenta and helicities of the scattering matrix elements as implemented in the portable partonic event generator pepper, we find substantial efficiency improvements in the studied processes. We focus our study on the highest final-state multiplicities in Drell-Yan and top-antitop pair production used in simulated samples for the Large Hadron Collider, which computationally are the most relevant ones. We find that the unweighting efficiencies improve by factors of 184 and 25, respectively, when compared to the standard approach of using a vegas-based optimization. We also compare continuous normalizing flows trained with flow matching against the previously studied normalizing flows based on coupling layers and find that the former leads to better results, faster training and a better scaling behavior across the studied multiplicity range, while the latter evaluate faster. When combining the advantages of both methods using the regflow approach, we find parton-level unweighted event generation walltime gains of about a factor of 10 at the highest final-state multiplicities.

Bothmann, E. [CERN; Gottingen U.] (ORCID:000000016↗

DOE OSTI2026

Trajectory-independent speed limits for controlled open quantum systems

Existing quantum speed limits for controlled open quantum systems depend on the specified trajectory. For example, lower bounds on quantum annealing times in the presence of dissipation depend explicitly on the chosen annealing schedule. Recently, schedule-independent speed limits have been derived for annealing in the closed quantum system setting [L. P. García-Pintos et al ., SciPost Phys. 18 , 159 (2025)]. In this work, we generalize these results to open quantum systems, deriving schedule-independent lower bounds for quantum annealing times in systems described by a Lindblad master equation. We analyze the interplay between coherent control and dissipation in single- and two-qubit examples, demonstrating that the derived lower bounds capture key scaling behavior with respect to the strength of the dissipator. Finally, we apply the bound to thermal state preparation and show that the bound matches the expected asymptotic behavior for an Ising model in the high-temperature limit.

open quantum systems & decoherence↗

DOE OSTI2020

Continuous phase transition between bosonic integer quantum Hall liquid and a trivial insulator: Evidence for deconfined quantum criticality

The deconfined quantum critical point, a prototype Landau-forbidden transition, could exist, in principle, in the phase transitions involving a symmetry-protected topological phase; however, examples of such kinds of transitions in physical systems are rare beyond one-dimensional systems. Here, using a density-matrix renormalization-group calculation, we unveil a bosonic integer quantum Hall phase in a two-dimensional correlated honeycomb lattice, by full identification of its internal structure from the topological K matrix. Furthermore, we demonstrate that imbalanced periodic chemical potentials can destroy the bosonic integer quantum Hall state and drive it into a featureless trivial (Mott) insulator, where all physical observables evolve smoothly across the critical point. At the critical point, the entanglement entropy reveals a characteristic scaling behavior, which is consistent with the critical field theory as an emergent QED3 with two flavors of Dirac fermions.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Characterizing random-singlet state in two-dimensional frustrated quantum magnets and implications for the double perovskite Sr 2 CuTe 1 – x W x O 6

Motivated by the experimental observation of a nonmagnetic phase in compounds with frustration and disorder, we study the ground state of a spin-1/2 square-lattice Heisenberg model with randomly distributed nearest-neighbor J 1 and next-nearest-neighbor J 2 couplings. By using the density matrix renormalization group (DMRG) calculation on a cylinder system with a circumference of up to ten lattice sites, we identify a disordered phase between the Néel and stripe magnetic phase with growing J 2 /J 1 in the presence of strong bond randomness. The vanished spin-freezing parameter indicates the absence of spin-glass order. The large-scale DMRG results unveil the size-scaling behaviors of the spin-freezing parameter, the power-law decay of the average spin correlation, and the exponential decay of the typical spin correlation, which all agree with the corresponding behavior in the one-dimensional random-singlet (RS) state and characterize the RS nature of this disordered phase. The DMRG simulation also provides insights and opportunities for characterizing a class of nonmagnetic states in two-dimensional frustrated magnets with disorder. Here, we also compare with existing experiments and suggest more measurements for understanding the spin-liquid-like behaviors in the double perovskite Sr 2 CuTe 1–x W x O 6 .

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2023

Suppression of one-dimensional weak localization by band asymmetry

Herein we investigate disorder-induced localization in metals that break time-reversal and inversion symmetries through their energy dispersion, ε k ≠ε– k , but lack Berry phases. In the perturbative regime of disorder, we show that weak localization is suppressed due to a mismatch of the Fermi velocities of left and right movers. To substantiate this analytical result, we perform quench numerics on chains shorter than the Anderson localization length ξ—the latter computed and verified to be finite using the recursive Green's function method—and find a sharp rise in the saturation value of the participation ratio due to band asymmetry, indicating a tendency to delocalize. Interestingly, for weak disorder strength η, we see a better fit to the scaling behavior ξ∝1/η 2 for asymmetric bands than conventional symmetric ones.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2024

Extended Kohler's rule of magnetoresistance in TaCo 2 Te 2

TaCo 2 Te 2 is recently reported to be an air-stable, high mobility van der Waals material with probable magnetic order. Here we investigate the scaling behavior of its magnetoresistance. Here we measured both the longitudinal (p xx ) and Hall (p xy ) magnetoresistivities of TaCo 2 Te 2 crystals in magnetic fields parallel to the c axis and found that the magnetoresistance violates the Kohler's rule MR similar to f [H/p 0 ] while obeying the extended Kohler's rule MR ~ f [H/(n T p 0 )], where MR ~ [p xx (H) - p 0 ]/p 0 , H is the magnetic field, n T is a thermal factor, and p xx (H) and p 0 are the resistivities at H and zero field, respectively. Further, while deviating from those of the densities of electrons (n e ) and holes (n h ) obtained from the two-band model analysis of the magnetoconductivities, the temperature dependence of n T is close to that of the Hall carrier densities n H calculated from the slopes of p xy (H) curves at low magnetic fields, providing a different way to obtain the thermal factor in the extended Kohler's rule.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Quantum criticality in the anisotropic spin-chain antiferromagnet Ca 2 ⁢CoH 2 ⁢(SeO 3 ) 4

Here, the quantum criticality of an XXZ-type spin chain antiferromagnet Ca 2 ⁢CoH 2 ⁢(SeO 3 ) 4 , subjected to an external magnetic field, was investigated through ultralow-temperature specific heat, magnetocaloric effect, and magnetic susceptibility measurements. In the absence of an external magnetic field, the system undergoes an antiferromagnetic phase transition at T N = 1.42K. The application of an external field gradually suppresses the antiferromagnetic order. Furthermore, an additional phase emerges in the intermediate field range when the zero-field antiferromagnetic order is suppressed, resulting in a quantum critical point at B s ≈ 2.55T. Near this field-induced quantum critical point, universal scaling behaviors are observed, characterized by the universality class parameters d=1, v=1/2, and z=2.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2023

Detailed study of quark-hadron duality in spin structure functions of the proton and neutron

Background: The response of hadrons, the bound states of the strong force (QCD), to external probes can be described in two different, complementary frameworks: as direct interactions with their fundamental constituents, quarks and gluons, or alternatively as elastic or inelastic coherent scattering that leaves the hadrons in their ground state or in one of their excited (resonance) states. The former picture emerges most clearly in hard processes with high momentum transfer, where the hadron response can be described by the perturbative expansion of QCD, while at lower energy and momentum transfers, the resonant excitations of the hadrons dominate the cross section. The overlap region between these two pictures, where both yield similar predictions, is referred to as quark-hadron duality and has been extensively studied in reactions involving unpolarized hadrons. Some limited information on this phenomenon also exists for polarized protons, deuterons, and 3 He nuclei, but not yet for neutrons. Purpose: In this paper, we present comprehensive and detailed results on the correspondence between the extrapolated deep inelastic structure function g 1 of both the proton and the neutron with the same quantity measured in the nucleon resonance region. Thanks to the fine binning and high precision of our data, and using a well-controlled perturbative QCD (pQCD) fit for the partonic prediction, we can make quantitative statements about the kinematic range of applicability of both local duality and global duality. Method: We use the most updated QCD global analysis results at high x from the Jefferson Lab Angular Momentum Collaboration to extrapolate the spin structure function g 1 into the nucleon resonance region and then integrate over various intervals in the scaling variable x. We compare the results with the large data set collected in the quark-hadron transition region by the CLAS Collaboration, including, for the first time, deconvoluted neutron data, integrated over the same intervals. We present this comparison as a function of the momentum transfer Q 2 . Results: We find that, depending on the integration interval and the minimum momentum transfer chosen, a clear transition to quark-hadron duality can be observed in both nucleon species. Furthermore, we show, for the first time, the approach to scaling behavior for g 1 measured in the resonance region at sufficiently high momentum transfer. Conclusions: Here, our results can be used to quantify the deviations from the applicability of pQCD for data taken at moderate energies and can help with extraction of quark distribution functions from such data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2024

Nonprompt direct-photon production in Au + Au collisions at $\sqrt{s_{NN}}$ = 200 GeV

The measurement of the direct-photon spectrum from Au + Au collisions at $\sqrt{s_{NN}}$ = 200 GeV is presented by the PHENIX Collaboration using the external-photon-conversion technique for 0%–93% central collisions in a transverse-momentum (𝑝 𝑇 ) range of 0.8–10 GeV/𝑐. An excess of direct photons, above prompt-photon production from hard-scattering processes, is observed for 𝑝 𝑇 < 6 GeV/𝑐. Nonprompt direct photons are measured by subtracting the prompt component, which is estimated as 𝑁 coll -scaled direct photons from 𝑝 + 𝑝 collisions at 200 GeV, from the direct-photon spectrum. Results are obtained for 0.8 < 𝑝 𝑇 < 6.0 GeV/𝑐 and suggest that the spectrum has an increasing inverse slope from ≈0.2 to 0.4 GeV/𝑐 with increasing 𝑝 𝑇 , which indicates a possible sensitivity of the measurement to photons from earlier stages of the evolution of the collision. In addition, like the direct-photon production, the 𝑝 𝑇 -integrated nonprompt direct-photon yields also follow a power-law scaling behavior as a function of collision-system size. Finally, the exponent, 𝛼, for the nonprompt component is found to be consistent with 1.1 with no apparent 𝑝 𝑇 dependence.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2024

Dark matter isocurvature from curvature

Isocurvature fluctuations, where the relative number density of particle species spatially varies, can be generated from initially adiabatic, or curvature, fluctuations if the various species fall out of or were never in thermal equilibrium. The freezing of the thermal relic dark matter abundance is one such case, but for modes that are still outside the horizon the amplitude is highly suppressed and originates from the small change in the local expansion rate due to the local space curvature produced by the curvature fluctuation. We establish a simple separate-universe method for calculating this generation that applies to both freeze-in and freeze-out models, identify three critical epochs for this process, and give general scaling behaviors for the amplitude in each case: the freezing epoch, the kinetic decoupling epoch and matter-radiation equality. Freeze-out models are typically dominated by spatially modulated annihilation from the latter epochs and can generate much larger isocurvature fluctuations compared with typical freeze-in models, albeit still very small and observationally allowed by cosmic microwave background measurements. Here, we illustrate these results with concrete models where the dark matter interactions are vector or scalar mediated.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2020

Reversal of Simple Hydrogenic Isotope Scaling Laws in Tokamak Edge Turbulence

The role of nonadiabatic electrons in regulating the hydrogenic isotope-mass scaling of gyrokinetic turbulence in tokamak fusion plasmas is assessed in the transition from ion-dominated core transport regimes to electron-dominated edge transport regimes. We propose a new isotope-mass scaling law that describes the electron-to-ion mass-ratio dependence of turbulent ion and electron energy fluxes. Here, the massratio dependence arises from the nonadiabatic response associated with fast electron parallel motion and plays a key role in altering—and in the case of the DIII-D edge, favorably reversing—the naive gyro-Bohm scaling behavior. In the reversed regime hydrogen energy fluxes are larger than deuterium fluxes, which is the opposite of the naive prediction.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2023

Beam Energy Dependence of Triton Production and Yield Ratio $N_t \times N_p / N^2_d$ in Au+Au Collisions at RHIC

We report the triton (t) production in midrapidity (|y| < 0.5) Au + Au collisions at $\sqrt{s_{NN}} = 7.7-200$ GeV measured by the STAR experiment from the first phase of the beam energy scan at the Relativistic Heavy Ion Collider. The nuclear compound yield ratio ($N_t \times N_p / N^2_d$), which is predicted to be sensitive to the fluctuation of local neutron density, is observed to decrease monotonically with increasing charged-particle multiplicity ($dN_{ch}/dη$) and follows a scaling behavior. The $dN_{ch}/dη$ dependence of the yield ratio is compared to calculations from coalescence and thermal models. Enhancements in the yield ratios relative to the coalescence baseline are observed in the 0%-10% most central collisions at 19.6 and 27 GeV, with a significance of 2.3σ and 3.4σ, respectively, giving a combined significance of 4.1σ. The enhancements are not observed in peripheral collisions or model calculations without critical fluctuation, and decreases with a smaller $p_T$ acceptance. The physics implications of these results on the QCD phase structure and the production mechanism of light nuclei in heavy-ion collisions are discussed.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2025

Spin fluctuations, absence of magnetic order, and crystal electric field studies in the Yb 3+ -based triangular lattice antiferromagnet Rb 3 ⁢Yb⁢(VO 4 ) 2

Here, we report a comprehensive experimental investigation of the structural, thermodynamic, static, and dynamic properties of a triangular lattice antiferromagnet Rb 3 ⁢Yb(VO 4 ) 2 . Through the analysis of magnetic susceptibility, magnetization, and specific heat, complemented by crystal electric field (CEF) calculations, we confirm the Kramers' doublet with effective spin 𝐽 eff = 1/2 ground state. Magnetic susceptibility and isothermal magnetization analysis reveal a weak antiferromagnetic interaction among the 𝐽 eff = 1/2 spins, characterized by a small Curie-Weiss temperature (𝜃$^{\textrm{LT}}_{\textrm{CW}}$ ≃−0.26 K) or a reduced exchange coupling (𝐽/𝑘 B ≃ 0.18 K). The 51 V NMR spectra and spin-lattice relaxation rate (1/𝑇 1 ) show no evidence of magnetic long-range-order down to 1.6 K but reflect strong influence of CEF excitations in the intermediate temperatures. At low temperatures, 1/𝑇 1 ⁡(𝑇) shows pronounced frequency dependence and 1/𝑇 1 vs field at different temperatures follows the scaling behavior, highlighting the role of paramagnetic fluctuations. The CEF calculations using the point charge approximation divulge a large energy gap ( ∼18.61 meV) between the lowest and second lowest energy doublets, further establishing Kramers' doublet as the ground state. Our calculations also reproduce the experimental magnetization and specific heat data and indicate an in-plane magnetic anisotropy. These findings position Rb 3⁢ Yb(VO 4 ) 2 as an ideal candidate to explore intrinsic quantum fluctuations and possible quantum spin-liquid physics in a Yb 3+ -based triangular lattice antiferromagnet.

Sebastian, Sebin J. [Indian Institute of Science E↗

DOE OSTI2026

Correlation-aware binning for small-angle neutron scattering via Gaussian-process inference

Binning in small-angle neutron scattering (SANS) is typically performed empirically, with fixed parameters chosen for convenience rather than statistical optimality. Such practices often fail to balance statistical precision and spatial resolution, leading to inconsistencies across instruments and datasets. Here we establish a correlation-aware framework that determines the optimal bin width from first principles by extending the classical Freedman–Diaconis (FD) rule to account for inter-bin correlations with a Gaussian process. In this formulation, the scattering intensity is treated as a smooth stochastic field whose statistical coherence is described by a covariance matrix. Analytical expressions of errors derived from this model yield closed-form criteria that separate the total deviation into contributions from counting noise, aliasing distortion and curvature-dependent correlation effects. Expressed in reduced variables, the resulting dimensionless error surface reveals a continuous transition from the uncorrelated FD regime to the correlation-dominated limit, providing a unified description of noise suppression and resolution control. Because the formulation depends only on the profile characteristics of scattering intensity I(Q), specifically its average intensity and first- and second-order derivatives, it applies generally to any SANS measurement regardless of sample, instrument or geometry. Experimental validation using small- and ultra-small-angle neutron scattering data confirms the predicted scaling behavior, demonstrating that correlation-aware inference systematically reduces mean-squared error and enables information-efficient reproducible data reduction across materials and instruments.

Tung, Chi-Huan [ORNL] (ORCID:0000000221972074)↗

DOE OSTI2024

Toward a Holistic Performance Evaluation of Large Language Models Across Diverse AI Accelerators

Artificial intelligence (AI) methods have become critical in scientific applications to help accelerate scientific discovery. Large language models (LLMs) are being considered a promising approach to address some challenging problems because of their superior generalization capabilities across domains. The effectiveness of the models and the accuracy of the applications are contingent upon their efficient execution on the underlying hardware infrastructure. Specialized Al accelerator hardware systems have recently become available for accelerating Al applications. However, the comparative performance of these AI accelerators on large language models has not been previously studied. In this paper, we systematically study LLMs on multiple AI accelerators and GPUs and evaluate their performance characteristics for these models. We evaluate these systems with (i) a micro-benchmark using a core transformer block, (ii) a GPT-2 model, and (iii) an 1,I,M-driven science use case, GenSLM. We present our findings and analyses of the models' performance to better understand the intrinsic capabilities of AI accelerators. Furthermore, our analysis takes into account key factors such as sequence lengths, scaling behavior, and sensitivity to gradient accumulation steps.

Emani, Murali↗

DOE OSTI2023

ExTreeM: Scalable Augmented Merge Tree Computation via Extremum Graphs

Over the last decade merge trees have been proven to support a plethora of visualization and analysis tasks since they effectively abstract complex datasets. Here, this paper describes the ExTreeM-Algorithm: A scalable algorithm for the computation of merge trees via extremum graphs. The core idea of ExTreeM is to first derive the extremum graph G of an input scalar field f defined on a cell complex K, and subsequently compute the unaugmented merge tree of f on G instead of K; which are equivalent. Any merge tree algorithm can be carried out significantly faster on G, since K in general contains substantially more cells than G. To further speed up computation, ExTreeM includes a tailored procedure to derive merge trees of extremum graphs. The computation of the fully augmented merge tree, i.e., a merge tree domain segmentation of K, can then be performed in an optional post-processing step. All steps of ExTreeM consist of procedures with high parallel efficiency, and we provide a formal proof of its correctness. Our experiments, performed on publicly available datasets, report a speedup of up to one order of magnitude over the state-of-the-art algorithms included in the TTK and VTK-m software libraries, while also requiring significantly less memory and exhibiting excellent scaling behavior.

97 MATHEMATICS AND COMPUTING↗