arXiv subjects

Gang Chen

Publications and source records attributed to Gang Chen.

At least 397 records · Page 22Linked to original sources

arXiv2021

Semi-discrete and fully discrete HDG methods for Burgers' equation

This paper proposes semi-discrete and fully discrete hybridizable discontinuous Galerkin (HDG) methods for the Burgers' equation in two and three dimensions. In the spatial discretization, we use piecewise polynomials of degrees $ k \ (k \geq 1), k-1$ and $ l \ (l=k-1; k) $ to approximate the scalar function, flux variable and the interface trace of scalar function, respectively. In the full discretization method, we apply a backward Euler scheme for the temporal discretization. Optimal a priori error estimates are derived. Numerical experiments are presented to support the theoretical results.

math.NA↗

arXiv2021

C-for-Metal: High Performance SIMD Programming on Intel GPUs

The SIMT execution model is commonly used for general GPU development. CUDA and OpenCL developers write scalar code that is implicitly parallelized by compiler and hardware. On Intel GPUs, however, this abstraction has profound performance implications as the underlying ISA is SIMD and important hardware capabilities cannot be fully utilized. To close this performance gap we introduce C-For-Metal (CM), an explicit SIMD programming framework designed to deliver close-to-the-metal performance on Intel GPUs. The CM programming language and its vector/matrix types provide an intuitive interface to exploit the underlying hardware features, allowing fine-grained register management, SIMD size control and cross-lane data sharing. Experimental results show that CM applications from different domains outperform the best-known SIMT-based OpenCL implementations, achieving up to 2.7x speedup on the latest Intel GPU.

cs.DC↗

arXiv2021

Blockchains vs. Distributed Databases: Dichotomy and Fusion

Blockchain has come a long way: a system that was initially proposed specifically for cryptocurrencies is now being adapted and adopted as a general-purpose transactional system. As blockchain evolves into another data management system, the natural question is how it compares against distributed database systems. Existing works on this comparison focus on high-level properties, such as security and throughput. They stop short of showing how the underlying design choices contribute to the overall differences. Our work fills this important gap and provides a principled framework for analyzing the emerging trend of blockchain-database fusion. We perform a twin study of blockchains and distributed database systems as two types of transactional systems. We propose a taxonomy that illustrates the dichotomy across four dimensions, namely replication, concurrency, storage, and sharding. Within each dimension, we discuss how the design choices are driven by two goals: security for blockchains, and performance for distributed databases. To expose the impact of different design choices on the overall performance, we conduct an in-depth performance analysis of two blockchains, namely Quorum and Hyperledger Fabric, and two distributed databases, namely TiDB, and etcd. Lastly, we propose a framework for back-of-the-envelope performance forecast of blockchain-database hybrids.

cs.DB↗

arXiv2021

A study of $Λ$ and $\barΛ$ polarization splitting by meson field in PICR hydrodynamic model

With the PICR hydrodynamic model, we study the polarization splitting between $Λ$ and $\barΛ$ at RHIC BES energy range, based on the meson field mechanism. Our results fit to the experimental data fairly well. Besides, two unexpected effect emerges: (1) the baryon density gradient has non-trivial and negative contribution to the polarization splitting; (2) for 7.7 GeV Au+Au collisions within the centrality range of 20\%-50\%, the polarization splitting surprisingly increases with the centrality decreases. The second effect might help to explain the significant signal of polarization splitting measured in STAR's Au+Au 7.7 Gev collisions.

hep-ph↗

arXiv2020

Neural Machine Translation: A Review of Methods, Resources, and Tools

Machine translation (MT) is an important sub-field of natural language processing that aims to translate natural languages using computers. In recent years, end-to-end neural machine translation (NMT) has achieved great success and has become the new mainstream method in practical MT systems. In this article, we first provide a broad review of the methods for NMT and focus on methods relating to architectures, decoding, and data augmentation. Then we summarize the resources and tools that are useful for researchers. Finally, we conclude with a discussion of possible future research directions.

cs.CL↗

arXiv2020

Electron quasi-itinerancy intertwined with quantum order by disorder in pyrochlore iridate magnetism

We point out the emergence of magnetism from the interplay of electron quasi-itinerancy and quantum order by disorder in pyrochlore iridates. Like other Mott insulating iridates, the Ir$^{4+}$ ion in pyrochlore iridates develops an effective ${J=1/2}$ moment from the on-site spin-orbit coupling. We consider the generic symmetry-allowed exchange between these local moments on a pyrochlore lattice and obtain the mean-field phase diagram. Assuming the superexchange is mediated by direct and/or indirect electron hopping via intermediate oxygens, we derive the exchange interactions in the strong coupling regime from the Hubbard model. This exchange has a degenerate classical ground state manifold and quantum fluctuation selects a non-coplanar ground state, known as quantum order by disorder. Extending to the intermediate coupling regime, the same non-coplanar order is selected from the degenerate manifold by the kinetic energy, which is dubbed "electron quasi-itinerancy". We discuss the experimental relevance of our results and electron quasi-itinerancy among other iridates and $4d/5d$ magnets.

cond-mat.str-el↗

arXiv2020

Dilemma in strongly correlated materials: Hund's metal vs relativistic Mott insulator

We point out the generic competition between the Hund's coupling and the spin-orbit coupling in correlated materials, and this competition leads to an electronic dilemma between the Hund's metal and the relativistic insulators. Hund's metals refer to the fate of the would-be insulators where the Hund's coupling suppresses the correlation and drives the systems into correlated metals. Relativistic Mott insulators refer to the fate of the would-be metals where the relativistic spin-orbit coupling enhances the correlation and drives the systems into Mott insulators. These contradictory trends are naturally present in many correlated materials. We study the competition between Hund's coupling and spin-orbit coupling in correlated materials and explore the interplay and the balance from these two contradictory trends. The system can become a spin-orbit-coupled Hund's metal or a Hund's assisted relativistic Mott insulator. Our observation could find a broad application and relevance to many correlated materials with multiple orbitals.

cond-mat.str-el↗

arXiv2020

Study of nuclear modification factors of (anti-)hadrons and light (anti-)nuclei in Pb-Pb collisions at $\sqrt{s_{\rm{NN}}}$ = 2.76 TeV

The nuclear modification factors ($R_{AA}$) of $π^{\pm}, p(\bar p)$, and $d(\bar d)$ with $|y|<0.5, p_T<20.0$~GeV/c in peripheral (40-60\%) and central (0-5\%) Pb-Pb collisions at $\sqrt {s_{NN}}=2.76$ TeV have been studied using the parton and hadron cascade ({\footnotesize PACIAE}) model plus the dynamically constrained phase space coalescence ({\footnotesize DCPC}) model. It is found that the $R_{AA}$ of light (anti)nuclei ($d, \bar d$) is similar to that of hadrons ($π^\pm, p, \bar p$), and the $R_{AA}$ of antiparticles is the same as that of particles. The suppression of $R_{AA}$ at high-$p_T$ strongly depends on event centrality and mass of the particles, i.e., the central collision is more suppressed than the peripheral collision. Besides, the yield ratios and double ratios for different particle species in $pp$ and Pb-Pb collisions are discussed, respectively. It is observed that the yield ratios and double ratios of $d$ to $p$ and $p$ to $π$ are similar to those of their anti-particles in three different collision systems, suggesting that the suppressions of matter ($π^{+}, p, d$) and the corresponding antimatter ($π^{-},\bar{p},\bar{d}$) are around the same level.

hep-ph↗

arXiv2020

LINDT: Tackling Negative Federated Learning with Local Adaptation

Federated Learning (FL) is a promising distributed learning paradigm, which allows a number of data owners (also called clients) to collaboratively learn a shared model without disclosing each client's data. However, FL may fail to proceed properly, amid a state that we call negative federated learning (NFL). This paper addresses the problem of negative federated learning. We formulate a rigorous definition of NFL and analyze its essential cause. We propose a novel framework called LINDT for tackling NFL in run-time. The framework can potentially work with any neural-network-based FL systems for NFL detection and recovery. Specifically, we introduce a metric for detecting NFL from the server. On occasion of NFL recovery, the framework makes adaptation to the federated model on each client's local data by learning a Layer-wise Intertwined Dual-model. Experiment results show that the proposed approach can significantly improve the performance of FL on local data in various scenarios of NFL.

cs.LG↗

arXiv2020

Emergent Halperin-Saslow mode and Gauge Glass in quantum Ising magnet TmMgGaO$_4$

We propose quenched disorders could bring novel quantum excitations and models to certain quantum magnets. Motivated by the recent experiments on the quantum Ising magnet TmMgGaO$_4$, we explore the effects of the quenched disorder and the interlayer coupling in this triangular lattice Ising antiferromagnet. It is pointed out that the weak quenched (non-magnetic) disorder would convert the emergent 2D Berezinskii-Kosterlitz-Thouless (BKT) phase and the critical region into a gauge glass. There will be an emergent Halperin-Saslow mode associated with this gauge glass. Using the Imry-Ma argument, we further explain the fate of the finite-field $C_3$ symmetry breaking transition at the low temperatures. The ferromagnetic interlayer coupling would suppress the BKT phase and generate a tiny ferromagnetism. With the quenched disorders, this interlayer coupling changes the 2D gauge glass into a 3D gauge glass, and the Halperin-Saslow mode persists. This work merely focuses on addressing a phase regime in terms of emergent U(1) gauge glass behaviors and hope to inspire future works and thoughts in weakly disordered frustrated magnets in general.

cond-mat.str-el↗

arXiv2020

BERT-JAM: Boosting BERT-Enhanced Neural Machine Translation with Joint Attention

BERT-enhanced neural machine translation (NMT) aims at leveraging BERT-encoded representations for translation tasks. A recently proposed approach uses attention mechanisms to fuse Transformer's encoder and decoder layers with BERT's last-layer representation and shows enhanced performance. However, their method doesn't allow for the flexible distribution of attention between the BERT representation and the encoder/decoder representation. In this work, we propose a novel BERT-enhanced NMT model called BERT-JAM which improves upon existing models from two aspects: 1) BERT-JAM uses joint-attention modules to allow the encoder/decoder layers to dynamically allocate attention between different representations, and 2) BERT-JAM allows the encoder/decoder layers to make use of BERT's intermediate representations by composing them using a gated linear unit (GLU). We train BERT-JAM with a novel three-phase optimization strategy that progressively unfreezes different components of BERT-JAM. Our experiments show that BERT-JAM achieves SOTA BLEU scores on multiple translation tasks.

cs.CL↗

arXiv2020

Some experimental schemes to identify quantum spin liquids

Despite the apparent ubiquity and variety of quantum spin liquids in theory, experimental confirmation of spin liquids remains to be a huge challenge. Motivated by the recent surge of evidences for spin liquids in a series of candidate materials, we highlight the experimental schemes, involving the thermal Hall transport and spectrum measurements, that can result in smoking-gun signatures of spin liquids beyond the usual ones. For clarity, we investigate the square lattice spin liquids and theoretically predict the possible phenomena that may emerge in the corresponding spin liquids candidates. The mechanisms for these signatures can be traced back to either the intrinsic characters of spin liquids or the external field-driven behaviors. Our conclusion does not depend on the geometry of lattices and can broadly apply to other relevant spin liquids.

cond-mat.str-el↗

arXiv2020

Clusterization transition between cluster Mott insulators on a breathing Kagomé lattice

Motivated by recent experimental progress on various cluster Mott insulators, we study an extended Hubbard model on a breathing Kagomé lattice with a single electron orbital and $1/6$ electron filling. Two distinct types of cluster localization are found in the cluster Mott regime due to the presence of the electron repulsion between neighboring sites, rather than from the on-site Hubbard interaction in the conventional Mott insulators. We introduce a unified parton construction framework to accommodate both type of cluster Mott insulating phase as well as a trivial Ferm liquid metal and discuss the phase transitions in the phase diagram. It is shown that, in one of the cluster localization phases, the strong inter-site repulsion results into locally metallic behavior within one of two triangular clusters on the breathing Kagomé lattice. We further comment on experimental relevance to existing Mo-based cluster magnets.

cond-mat.str-el↗

arXiv2020

Towards a qualitative hadron-parton correspondence in the nuclear modification factor

In this work, we propose a method to show the correspondence between hadron and its quark component nuclear modification factors. A parton and hadron cascade model PACIAE based on the PYTHIA6.4 is employed to calculate the hadron and its quark component nuclear modification factors in the 0-5\% most central $Pb+Pb$ collisions at $\sqrt{s_{NN}}$=2.76 TeV. It turns out that the hadron nuclear modification factor is usually smaller than that of its quark component. On the other hand, it is shown in our study that the "dead cone effect" is more likely to be identified with the quarks and mesons but not with the baryon states obviously.

nucl-th↗

arXiv2020

Timing analysis of the black hole candidate EXO 1846-031 with Insight-HXMT monitoring

We present the observational results from a detailed timing analysis of the black hole candidate EXO 1846-031 during its outburst in 2019 with the observations of Insight-HXMT, NICER and MAXI. This outburst can be classfied roughly into four different states. Type-C quasi-periodic oscillations (QPOs) observed by NICER (about 0.1-6Hz) and Insight-HXMT (about 0.7-8Hz) are also reported in this work. Meanwhile, we study various physical quantities related to QPO frequency.The QPO rms-frequency relationship in three energy band 1-10 keV indicates that there is a turning pointing in frequency around 2 Hz,which is similar to that of GRS 1915+105. A possible hypothesis for the relationship above may be related to the inclination of the source, which may require a high inclination to explain it. The relationships between QPO frequency and QPO rms,hardness,total fractional rms and count rate have also been found in other transient sources, which can indicate that the origin of type-C QPOs is non-thermal.

astro-ph.HE↗

arXiv2020

Discovery of oscillations above 200 keV in a black hole X-ray binary with Insight-HXMT

Low-frequency quasi-periodic oscillations (LFQPOs) are commonly found in black hole X-ray binaries, and their origin is still under debate. The properties of LFQPOs at high energies (above 30 keV) are closely related to the nature of the accretion flow in the innermost regions, and thus play a crucial role in critically testing various theoretical models. The Hard X-ray Modulation Telescope (Insight-HXMT) is capable of detecting emissions above 30 keV, and is therefore an ideal instrument to do so. Here we report the discovery of LFQPOs above 200 keV in the new black hole MAXI J1820+070 in the X-ray hard state, which allows us to understand the behaviours of LFQPOs at hundreds of kiloelectronvolts. The phase lag of the LFQPO is constant around zero below 30 keV, and becomes a soft lag (that is, the high-energy photons arrive first) above 30 keV. The soft lag gradually increases with energy and reaches ~0.9s in the 150-200 keV band. The detection at energies above 200 keV, the large soft lag and the energy-related behaviors of the LFQPO pose a great challenge for most currently existing models, but suggest that the LFQPO probably originates from the precession of a small-scale jet.

astro-ph.HE↗

arXiv2020

Schur Rings over Z \times Z_3

For the direct product $\cZ\times \cZ_3$ of infinite cyclic group $\cZ$ and a cyclic group $\cZ_3$ of order $3$, the schur rings over it are classified. In particular, all the schur rings are proved to be traditional.

math.GR↗

arXiv2020

Superconvergent Interpolatory HDG methods for reaction diffusion equations II: HHO-inspired methods

In J. Sci. Comput., 81: 2188-2212, 2019, we considered a superconvergent hybridizable discontinuous Galerkin (HDG) method, defined on simplicial meshes, for scalar reaction diffusion equations and showed how to define an interpolatory version which maintained its convergence properties. The interpolatory approach uses a locally postprocessed approximate solution to evaluate the nonlinear term, and assembles all HDG matrices once before the time intergration leading to a reduction in computational cost. The resulting method displays a superconvergent rate for the solution for polynomial degree $k\ge 1$. In this work, we take advantage of the link found between the HDG and the hybrid high-order (HHO) methods, in ESAIM Math. Model. Numer. Anal., 50: 635-650, 2016 and extend this idea to the new, HHO-inspired HDG methods, defined on meshes made of general polyhedral elements, uncovered therein}. We prove that the resulting interpolatory HDG methods converge at the same rate as for the linear elliptic problems. Hence, we obtain superconvergent methods for $k\ge 0$ by some methods. We present numerical results to illustrate the convergence theory.

math.NA↗