arXiv subjects

Ran Guo

Publications and source records attributed to Ran Guo.

At least 19 recordsLinked to original sources

arXiv2026

Simulations of self-accelerating electron phase space holes in an applied electric field

The self-acceleration of electron phase space holes in an applied electric field is investigated via one-dimensional electrostatic Vlasov simulations. The electron holes (EHs) are initialized in a self-consistent manner with immobile ions, and the ion response is enabled at the beginning of simulations. A benchmark simulation is conducted to confirm the EH self-acceleration in the absence of the external electric field. Then, we investigate the EH behaviors by applying the uniform and sinusoidal electric fields, respectively. The effects of different strengths and durations of these external electric fields are studied. It is found that the uniform electric field applied in the direction of the self-acceleration can delay the onset of this process and change the final speed of EHs. The applied sinusoidal electric field can fix the EHs at their initial positions and suppress the self-acceleration if the electric field amplitude and duration are appropriate. In addition, it is observed that these external electric fields can induce the splitting of EHs and the generation of secondary EHs. The physical mechanisms of these phenomena are discussed in detail.

physics.plasm-ph↗

arXiv2026

Enhancement of $J$$_c$ by Proton Irradiation in HgBa$_2$Ca$_2$Cu$_3$O$_8$$_+$$_\delta$ Single Crystals

Critical current density is the key parameter for the practical application of superconductivity. In this study, 3 MeV proton irradiation experiments were conducted on HgBa$_2$Ca$_2$Cu$_3$O$_8$$_+$$_\delta$ single crystals to introduce pinning centers. The critical current density is found to be strongly enhanced after the irradiation with its maximum at a dose of 1$\times$10$^{16}$/cm$^2$, where the self-field critical current density at 2 K is enhanced from 5.5 MA/cm$^2$ to 26 MA/cm$^2$. At 77 K, the self-field critical current density for all irradiated crystals is over 0.1 MA/cm$^2$. The power-law dependence of the critical current density on the magnetic field is observed after irradiation, with a large power-law exponent $\alpha$ close to 1. A monotonic magnetic field dependence of the normalized magnetic relaxation rate is observed, which could be attributed to the low irreversibility field caused by the large anisotropy in Hg1223 single crystals. Through the analysis of the pinning force density of the crystal before and after irradiation, a clear mechanism change has been observed.

cond-mat.supr-con↗

arXiv2025

Kinetic Alfv\'en waves in the temperature anisotropic space plasma with a kappa-Maxwellian distribution

The dispersion and damping rate of kinetic Alfv\'en waves are studied in temperature anisotropic space plasma with kappa-Maxwellian distribution. Employing a kinetic approach, the wave frequency and damping rate of kinetic Alfv\'en waves and the modified ion acoustic waves are derived in a low \b{eta} plasma, which both depend on the parameters \k{appa} and . The numerical analyses show that the wave frequency of kinetic Alfv\'en waves is larger in kappa-Maxwellian plasma than that in Maxwellian case. The wave frequency of the modified ion acoustic waves in kappa-Maxwellian plasma is larger in the short-wave region but smaller in the long-wave region than that in Maxwellian case. Again, we found that the damping rate of kinetic Alfv\'en waves in kappa-Maxwellian plasma is stronger than that in Maxwellian case. The damping rate of modified ion acoustic waves in kappa-Maxwellian plasma is stronger in the short-wave region but weaker in the long-wave region than that in Maxwellian case. The impact of the parameter on the two modes is relatively small because we consider the low \b{eta} case. These results are helpful for us to understand better the characteristics of kinetic Alfv\'en waves in space plasma.

physics.plasm-ph↗

arXiv2025

Increase of critical current density in FeSe superconductor by strain effect

Conventional $J_c$-enhancement methods like doping and irradiation often introduce extrinsic elements or defects, altering intrinsic properties. Here, we report a significant $J_c$ enhancement in FeSe single crystals through compressive strain applied using a glass-fiber-reinforced plastic substrate with anisotropic thermal contraction during cooling. Under zero field at 2 K, $J_{\text{c}}$ increases by a factor of $\sim$4 from $\sim 2.3 \times 10^{4}$ to $\sim 8.7 \times 10^{4}$ A cm$^{-2}$; at 5 T, it achieves an order-of-magnitude enhancement, rising from $\sim 1.0 \times 10^{3}$ to $\sim 1.0 \times 10^{4}$ A cm$^{-2}$. Analysis based on the Dew-Hughes model of the $f_{\text{p}}$(h) relationship shows that strain strengthens vortex pinning, and shifts the pinning mechanism from point-like pinning to combined point and surface pinnings. This work offers an effective method to enhance FeSe's current-carrying limitation, deepens understanding of iron-based superconductors' pinning mechanisms, and highlights strain engineering's potential for optimizing superconducting performance.

cond-mat.supr-con↗

arXiv2025

UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context Learning

While Mixture of Experts (MoE) models achieve remarkable efficiency by activating only subsets of parameters, they suffer from high memory access costs during inference. Memory-layer architectures offer an appealing alternative with very few memory access, but previous attempts like UltraMem have only matched the performance of 2-expert MoE models, falling significantly short of state-of-the-art 8-expert configurations. We present UltraMemV2, a redesigned memory-layer architecture that closes this performance gap. Our approach introduces five key improvements: integrating memory layers into every transformer block, simplifying value expansion with single linear projections, adopting FFN-based value processing from PEER, implementing principled parameter initialization, and rebalancing memory-to-FFN computation ratios. Through extensive evaluation, we demonstrate that UltraMemV2 achieves performance parity with 8-expert MoE models under same computation and parameters but significantly low memory access. Notably, UltraMemV2 shows superior performance on memory-intensive tasks, with improvements of +1.6 points on long-context memorization, +6.2 points on multi-round memorization, and +7.9 points on in-context learning. We validate our approach at scale with models up to 2.5B activated parameters from 120B total parameters, and establish that activation density has greater impact on performance than total sparse parameter count. Our work brings memory-layer architectures to performance parity with state-of-the-art MoE models, presenting a compelling alternative for efficient sparse computation.

cs.LG↗

arXiv2025

Investigations of the kinetic ion-acoustic soliton by the Bernstein-Greene-Kruskal integral method

The solitary waves are investigated through the Bernstein-Greene-Kruskal integral method with the ion response. We consider two specific cases of ions, i.e., the single stream with the waterbag distribution and the two counter-propagating streams with the Maxwellian distribution. The trapped electron distributions are derived for both two cases. The results show that the trapped electron distribution can be either a hole or a hump in the phase space, depending on the competition between the contributions from the passing electron distribution, the potential profile, and the ion response. We obtain the boundary between the ion-acoustic soliton and the electron hole in the parameter space. The effects of the potential amplitude, width, and the ion-to-electron mass ratio on the separatrices are discussed. The Vlasov simulations are conducted to verify the stability of the ion-acoustic soliton constructed by the integral method.

physics.plasm-ph↗

arXiv2024

Ultra-Sparse Memory Network

It is widely acknowledged that the performance of Transformer models is logarithmically related to their number of parameters and computational complexity. While approaches like Mixture of Experts (MoE) decouple parameter count from computational complexity, they still face challenges in inference due to high memory access costs. This work introduces UltraMem, incorporating large-scale, ultra-sparse memory layer to address these limitations. Our approach significantly reduces inference latency while maintaining model performance. We also investigate the scaling laws of this new architecture, demonstrating that it not only exhibits favorable scaling properties but outperforms MoE. In experiments, the largest UltraMem we train has 20 million memory slots. The results show that our method achieves state-of-the-art inference speed and model performance within a given computational budget, paving the way for billions of slots or experts.

cs.LG↗

arXiv2024

Drift instabilities driven by slab ion temperature gradient in suprathermal plasmas

The drift instabilities driven by the slab ion temperature gradient (ITG) in Kappa-distributed plasmas are investigated by the kinetic method. The linear dispersion relation is given in an integral representation involving only the standard plasma dispersion function. The wave frequency and growth rate are derived without the density inhomogeneity. Numerical solutions of the dispersion equation are conducted to show the different effects of the suprathermal ions and electrons. We find that the suprathermal ions can enhance the instability in large wavenumbers but suppress it in small wavenumbers. Thus, the suprathermalization of ions could be one of the factors leading to a lower limit of wavenumbers for the ITG instabilities. Besides, the numerical calculations also imply that the thermal speed ratio affects the intensities of the suprathermal effects. Finally, in the presence of density inhomogeneity, the ITG instability boundary is numerically analyzed.

physics.plasm-ph↗

arXiv2023

Different effects of suprathermal electrons and ions on drift instabilities in non-uniform plasmas

The paper studies the drift instability in Kappa-distributed inhomogeneous plasmas. Weak inhomogeneity and local approximation are assumed. The linear dispersion relation is expressed in a novel integral representation that involves the standard plasma dispersion functions defined in Maxwellian plasmas rather than the generalized plasma dispersion function modified by the Kappa distribution. The analytical solutions of wave frequency and growth rate are derived when the wave speed is larger than the ion thermal speed but smaller than the electron thermal speed. The unstable drift mode is found to have a lower limit of wavenumber due to the suprathermal electrons, which does not exist in Maxwellian plasmas. The population of suprathermal ions affects the critical wavenumber for instability but does not determine its emergence. Moreover, the wave frequency, growth rate, and critical wavenumber are numerically solved from the dispersion relation. The results suggest that the critical wavenumber cannot be neglected with the typical parameters of space plasmas.

physics.plasm-ph↗

arXiv2022

Small amplitude ion-acoustic solitary waves in a four-component magneto-rotating plasma with a modified Cairns-Tsallis distribution

The small amplitude ion-acoustic solitary waves in the magneto-rotating plasma consisting of cold fluid ions, hot positrons, and the two-temperature electrons (cold and hot electrons) are investigated when the electrons obey a modified Cairns-Tsallis distribution. By using the reductive perturbation method, we derive the Korteweg-de Vries equation and the modified Korteweg-de Vries equation and obtain the small amplitude ion-acoustic solitary wave solutions. The dependences of solitary wave solutions on the nonextensive q-parameter, the nonthermal alpha-parameter and the plasma physical quantities are analyzed numerically. We show the significant effects of the nonextensive q-parameter and the nonthermal alpha-parameter etc. on the ion-acoustic solitary waves.

physics.plasm-ph↗

arXiv2022

Effects of suprathermal electrons on electron-acoustic instabilities

We study the electron-acoustic instabilities in plasmas with two kappa-distributed electrons and stationary ions. The instabilities are driven by the relative drift between two electron components. The suprathermal effects of different species on growth rates and instability thresholds are analyzed and compared by numerical calculations. The present study reveals that the suprathermal electrons with slow most probable speed play more important roles than the suprathermal ones with fast most probable speed. The former significantly reduces the instabilities and raises the instability thresholds. The parameters used in this study are inspired from the observations in Earth's magnetopause.

physics.plasm-ph↗

arXiv2022

OneFlow: Redesign the Distributed Deep Learning Framework from Scratch

Deep learning frameworks such as TensorFlow and PyTorch provide a productive interface for expressing and training a deep neural network (DNN) model on a single device or using data parallelism. Still, they may not be flexible or efficient enough in training emerging large models on distributed devices, which require more sophisticated parallelism beyond data parallelism. Plugins or wrappers have been developed to strengthen these frameworks for model or pipeline parallelism, but they complicate the usage and implementation of distributed deep learning. Aiming at a simple, neat redesign of distributed deep learning frameworks for various parallelism paradigms, we present OneFlow, a novel distributed training framework based on an SBP (split, broadcast and partial-value) abstraction and the actor model. SBP enables much easier programming of data parallelism and model parallelism than existing frameworks, and the actor model provides a succinct runtime mechanism to manage the complex dependencies imposed by resource constraints, data movement and computation in distributed deep learning. We demonstrate the general applicability and efficiency of OneFlow for training various large DNN models with case studies and extensive experiments. The results show that OneFlow outperforms many well-known customized libraries built on top of the state-of-the-art frameworks. The code of OneFlow is available at: https://github.com/Oneflow-Inc/oneflow.

cs.DC↗

arXiv2022

Kinetic theory of $\mathrm{sech}^2x$ electron holes and applications to Kappa-distributed plasmas

The kinetic theory of $\mathrm{sech}^2 x$-type electron holes is studied. The potential of the electron holes is solved in the weak amplitude limit by the pseudo-potential method. We investigate the existence condition of the $\mathrm{sech}^2 x$ electron holes. It indicates that the derivatives of trapped and untrapped distributions at the separatrix play significant roles in determining the potential profile. The theory is then applied to the Kappa-distributed plasmas. The amplitude and width of the $\mathrm{sech}^2 x$ electron holes are analyzed. Finally, the theoretical results are verified by numerical calculations.

physics.plasm-ph↗

arXiv2021

The electron acoustic waves in plasmas with two kappa-distributed electrons at the same temperatures and immobile ions

The linear electron acoustic waves propagating in plasmas with two kappa-distributed electrons and stationary ions are investigated. The temperatures of the two electrons are assumed to be the same, but the kappa indices are not. It shows that if one kappa index is small enough and the other one is large enough, a weak damping regime of the electron acoustic waves exists. The dispersions and damping rates are numerically studied. The parameter spaces for the weakly damped electron acoustic waves are analyzed. Moreover, the electron acoustic waves in the present model are compared with those in other models, especially the plasmas with two-temperature electrons. At last, we perform Vlasov-Poisson simulations to verify the theory.

physics.plasm-ph↗

arXiv2020

Stationary states of polytropic plasmas

In this paper, we prove that the kappa distribution is the stationary solution of the Vlasov-Poisson system in an inhomogeneous plasma under the polytropic equation of state and an assumption restricting the local velocity distribution to a specific mathematical form. The profiles of density, temperature, and electric potential are obtained theoretically. The kappa index can be determined if the initial state is known. In order to verify the theory, particle-in-cell simulations are made and the results show excellent agreement with the theoretical predictions for density, temperature, and velocity distributions of electrons. It is shown that the electron velocity distribution of spatially inhomogeneous plasma evolves from an initial Maxwellian to the final kappa distribution. It is also found that the value of kappa index in the final stationary state depends on the initial state of plasma.

physics.plasm-ph↗

arXiv2020

An improved method to derive the kappa distribution in polytropic plasmas

This paper studies sufficient conditions for deriving the kappa distribution in polytropic plasmas by an improved method compared with the previous work [R. Guo, Phys. Plasmas \textbf{27}, 122104 (2020)]. We find that the polytropic equation of state can lead to the kappa distribution without any other assumptions in one dimension. In higher dimensions, an extra assumption that the global distribution must only depend on the energy is still needed. In addition, the self-consistence of the electrostatic potential is investigated in such plasmas. The study implies that in single-component plasmas, the potential is constrained by the Lane-Emden equation. However, in two-component plasmas, any arbitrary potential could exist with a suitable background ion density. Finally, we discuss the connections between the kappa distribution and the polytrope.

physics.plasm-ph↗

arXiv2019

Slowing down of charged particles in dusty plasmas with pawer-law kappa-distributions

We study the slowing down of particle beams passing through the dusty plasma with power-law kappa-distributions. Three plasma components, electrons, ions and dust particles, can have different kappa-parameter. We derive the deceleration factor (the velocity moment equation) and the slowing down time of a test particle, and numerically study the slowing down properties of an electron beam, a proton beam and a dust particle beam, respectively, in the kappa-distributed dusty plasma. We show that the slowing down properties of particle beams depend strongly on the kappa-parameters of the plasma components, and the dust component plays a dominant role in the slowing down. And the slowing down also depends on mass and charge of the dust particles in the dusty plasma. More detailed results are shown in 17 numerical graphs.

physics.plasm-ph↗

arXiv2018

The collisional relaxation rate of kappa-distributed plasma with multiple components

The kappa-distributed fully ionized plasma with collisional interaction is investigated. The Fokker-Planck equation with Rosenbluth potential is employed to describe such a physical system. The results show that the kappa distribution is not a stationary distribution unless the parameter kappa tends to infinity. The general expressions of collisional relaxation rate of multiple-component plasma with kappa distribution are derived and discussed in specific cases in details. For the purpose of visual illustration, we also give those results numerically in figures. All the results show that the parameter kappa plays a significant role in relaxation rate.

physics.plasm-ph↗