arXiv subjects

Lei Wang

Publications and source records attributed to Lei Wang.

At least 1,243 records · Page 69Linked to original sources

arXiv2014

BigDataBench: a Big Data Benchmark Suite from Internet Services

As architecture, systems, and data management communities pay greater attention to innovative big data systems and architectures, the pressure of benchmarking and evaluating these systems rises. Considering the broad use of big data systems, big data benchmarks must include diversity of data and workloads. Most of the state-of-the-art big data benchmarking efforts target evaluating specific types of applications or system software stacks, and hence they are not qualified for serving the purposes mentioned above. This paper presents our joint research efforts on this issue with several industrial partners. Our big data benchmark suite BigDataBench not only covers broad application scenarios, but also includes diverse and representative data sets. BigDataBench is publicly available from http://prof.ict.ac.cn/BigDataBench . Also, we comprehensively characterize 19 big data workloads included in BigDataBench with varying data inputs. On a typical state-of-practice processor, Intel Xeon E5645, we have the following observations: First, in comparison with the traditional benchmarks: including PARSEC, HPCC, and SPECCPU, big data applications have very low operation intensity; Second, the volume of data input has non-negligible impact on micro-architecture characteristics, which may impose challenges for simulation-based big data architecture research; Last but not least, corroborating the observations in CloudSuite and DCBench (which use smaller data inputs), we find that the numbers of L1 instruction cache misses per 1000 instructions of the big data applications are higher than in the traditional benchmarks; also, we find that L3 caches are effective for the big data applications, corroborating the observation in DCBench.

cs.DB↗

arXiv2014

Protecting sing-photon multi-mode W state from photon loss

Single-photon entanglement is of major importance in current quantum communications. However, it is sensitive to photon loss. In this paper, we discuss the protection of single-photon multi-mode W state with noiseless linear amplification. It is shown that the amplification factor is only decided with the transmission coefficient of the variable fiber beam splitters, and it does not change with the number of the spatial mode. This protocol may be useful in current quantum information processing.

quant-ph↗

arXiv2014

A Generalized Probabilistic Framework for Compact Codebook Creation

Compact and discriminative visual codebooks are preferred in many visual recognition tasks. In the literature, a number of works have taken the approach of hierarchically merging visual words of an initial large-sized codebook, but implemented this approach with different merging criteria. In this work, we propose a single probabilistic framework to unify these merging criteria, by identifying two key factors: the function used to model class-conditional distribution and the method used to estimate the distribution parameters. More importantly, by adopting new distribution functions and/or parameter estimation methods, our framework can readily produce a spectrum of novel merging criteria. Three of them are specifically focused in this work. In the first criterion, we adopt the multinomial distribution with Bayesian method; In the second criterion, we integrate Gaussian distribution with maximum likelihood parameter estimation. In the third criterion, which shows the best merging performance, we propose a max-margin-based parameter estimation method and apply it with multinomial distribution. Extensive experimental study is conducted to systematically analyse the performance of the above three criteria and compare them with existing ones. As demonstrated, the best criterion obtained in our framework achieves the overall best merging performance among the comparable merging criteria developed in the literature.

cs.CV↗

arXiv2014

Measurement of Collective Dynamical Mass of Dirac Fermions in Graphene

Individual electrons in graphene behave as massless quasiparticles. In surprising twist, it is inferred from plasmonic investigations that collectively excited graphene electrons must exhibit non-zero mass and its inertial acceleration is essential for graphene plasmonics. Despite such importance, this collective mass has defied direct unequivocal measurement. It may be directly measured by accelerating it with a time-varying voltage and quantifying the phase delay of the resulting current; this voltage-current phase relation would manifest as kinetic inductance, representing the collective inertia's reluctance to accelerate. However, at optical (infrared) frequencies phase measurement of current is generally difficult and at microwave frequencies the inertial phase delay has been buried under electron scattering. Here we directly, precisely measure the kinetic inductance, thus, collective mass, by combining innovative device engineering that reduces electron scattering and delicate microwave phase measurements. Particularly, encapsulation of graphene between hexagonal-boron-nitride layers, one-dimensional edge contacts, and a proximate top gate configured as microwave ground together enable resolving the inertial phase delay from the electron scattering. Beside the fundamental importance, the kinetic inductance demonstrated here to be orders-of-magnitude larger than magnetic inductance can dramatically miniaturize radio-frequency integrated circuits. Moreover, its bias-dependency heralds a solid-state voltage-controlled inductor to complement the prevalent voltage-controlled capacitor.

cond-mat.mes-hall↗

arXiv2013

Measuring the X-ray luminosities of SDSS DR7 clusters from RASS

We use ROSAT All Sky Survey (RASS) broadband X-ray images and the optical clusters identified from SDSS DR7 to estimate the X-ray luminosities around $\sim 65,000$ candidate clusters with masses $\ga 10^{13}\msunh$ based on an Optical to X-ray (OTX) code we develop. We obtain a catalogue with X-ray luminosity for each cluster. This catalog contains 817 clusters (473 at redshift $z\le 0.12$) with $S/N> 3$ in X-ray detection. We find about $65\%$ of these X-ray clusters have their most massive member located near the X-ray flux peak; for the rest $35\%$, the most massive galaxy is separated from the X-ray peak, with the separation following a distribution expected from a NFW profile. We investigate a number of correlations between the optical and X-ray properties of these X-ray clusters, and find that: the cluster X-ray luminosity is correlated with the stellar mass (luminosity) of the clusters, as well as with the stellar mass (luminosity) of the central galaxy and the mass of the halo, but the scatter in these correlations is large. Comparing the properties of X-ray clusters of similar halo masses but having different X-ray luminosities, we find that massive halos with masses $\ga 10^{14}\msunh$ contain a larger fraction of red satellite galaxies when they are brighter in X-ray. ... A cluster catalog containing the optical properties of member galaxies and the X-ray luminosity is available at {\it http://gax.shao.ac.cn/data/Group.html}.

astro-ph.CO↗

arXiv2013

Dark matter in little Higgs model under current experimental constraints from LHC, Planck and Xenon

We examine the status of dark matter (heavy photon) in the littlest Higgs model with T-parity (LHT) in light of the new results from the LHC Higgs search, the Planck dark matter relic density and the XENON100 limit on the dark matter scattering off the nucleon. We obtain the following observations: (i) For the LHC Higgs data, the LHT can well be consistent with the CMS results but disfavored by the ATLAS observation of diphoton enhancement; (ii) For the dark matter relic density, the heavy photon in the LHT can account for the Planck data for the small mass splitting of mirror lepton and heavy photon; (iii) For the dark matter scattering off the nucleon, the heavy photon can give a spin-independent cross section below the XENON100 upper limit for $m_{A_H}>95$ GeV ($f> 665$ GeV); (iv) A fit using the CMS Higgs data gives the lowest chi-square of 2.63 (the SM value is 4.75) at $f\simeq$ 1120 GeV and $m_{A_H}\simeq$ 170 GeV (at this point the dark matter constraints from Planck and XENON100 can also be satisfied). Such a best point and its nearby favored region (even for a $f$ value up to 3.8 TeV) can be covered by the future XENON1T (2017) experiment.

hep-ph↗

arXiv2013

Alignments of galaxies within cosmic filaments from SDSS DR7

Using a sample of galaxy groups selected from the Sloan Digital Sky Survey Data Release 7 (SDSS DR7), we examine the alignment between the orientation of galaxies and their surrounding large scale structure in the context of the cosmic web. The latter is quantified using the large-scale tidal field, reconstructed from the data using galaxy groups above a certain mass threshold. We find that the major axes of galaxies in filaments tend to be preferentially aligned with the directions of the filaments, while galaxies in sheets have their major axes preferentially aligned parallel to the plane of the sheets. The strength of this alignment signal is strongest for red, central galaxies, and in good agreement with that of dark matter halos in N-body simulations. This suggests that red, central galaxies are well aligned with their host halos, in quantitative agreement with previous studies based on the spatial distribution of satellite galaxies. There is a luminosity and mass dependence that brighter and more massive galaxies in filaments and sheets have stronger alignment signals. We also find that the orientation of galaxies is aligned with the eigenvector associated with the smallest eigenvalue of the tidal tensor. These observational results indicate that galaxy formation is affected by large-scale environments, and strongly suggests that galaxies are aligned with each other over scales comparable to those of sheets and filaments in the cosmic web.

astro-ph.CO↗

arXiv2013

Validity of Fourier's law in one-dimensional momentum-conserving lattices with asymmetric interparticle interactions

We have numerically studied heat conduction in a few one-dimensional momentum-conserving lattices with asymmetric interparticle interactions by the nonequilibrium heat bath method, the equilibrium Green-Kubo method, and the heat current power spectra analysis. Very strong finite-size effects are clearly observed. Such effects make the heat conduction obey a Fourier-like law in a wide range of lattice lengths. However, in yet longer lattice lengths, the heat conductivity regains its power-law divergence. Therefore the power-law divergence of the heat conductivity in the thermodynamic limit is verified, as is expected by many existing theories.

cond-mat.stat-mech↗

arXiv2013

Thermodynamics and magnetic properties of the anisotropic 3D Hubbard model

We study the 3D Hubbard model with anisotropic nearest neighbor tunneling amplitudes using the dynamical cluster approximation and compare the results with a quantum simulation experiment using ultracold fermions in an optical lattice, focussing on magnetic correlations. We find that the short-range spin correlations are significantly enhanced in the direction with stronger tunneling amplitudes. Our results agree with the experimental observations and show that the experimental temperature is lower than the strong tunneling amplitude. We characterize the system by examining the spin correlations beyond neighboring sites and determine the distribution of density, entropy and spin correlation in the trapped system. We furthermore investigate the dependence of the critical entropy at the Néel transition on anisotropy.

cond-mat.quant-gas↗

arXiv2013

Seeing Hofstadter's Butterfly in Atomic Fermi Gases

We propose a novel way to detect the fractal energy spectrum of the Hofstadter model from the density distributions of ultracold fermions in an external trap. At low temperature, the local compressibility is proportional to the density of states of the system which reveals the fractal energy spectrum. However, thermal broadening and noises in the real experimental situation inevitably smear out fine features in the density distribution. To overcome this difficulty, we use the maximum entropy method to extract the density of states directly from the noisy thermal density distributions. Simulations show that one is able to restore the core feature of the Hofstadter's butterfly spectrum with current experimental techniques. By further reducing the noise or the temperature, one can refine the resolution and observe fine structures of the butterfly spectrum.

cond-mat.quant-gas↗

arXiv2013

Splashing phenomena of room temperature liquid metal droplet striking on the pool of the same liquid under ambient air environment

In this article, the fluid dynamics of room temperature liquid metal (RTLM) droplet impacting onto a pool of the same liquid in ambient air was investigated. A series of experiments were conducted in order to disclose the influence of the oxidation effect on the impact dynamics. The droplet shape and impact phenomenology were recorded with the aid of a high-speed digital camera. The impact energy stored in the splash structures was estimated via a theoretical model and several morphological parameters obtained from instantaneous images of the splash. It was observed that the droplet shape and the splashing morphology of RTLM were drastically different from those of water, so was the impact dynamics between room temperature LM pool and high temperature LM pool. The energy analysis disclosed that the height of the jet is highly sensitive to the viscosity of the fluid, which is subjected to the oxidation effect and temperature effect simultaneously, and thus perfectly explained the phenomena. These basic findings are important for the application of RTLM in a series of newly emerging technologies such as liquid metal based spray cooling, ink-jet printed electronics, interface material painting and coating, metallurgy, and 3D packages, etc.

physics.flu-dyn↗

arXiv2013

Ferromagnetism of the Repulsive Atomic Fermi Gas: three-body recombination and domain formation

The simplest model for itinerant ferromagnetism, the Stoner model, has so far eluded experimental observation in repulsive ultracold fermions due to rapid three-body recombination at large scattering lengths. Here we show that a ferromagnetic phase can be stabilised by imposing a moderate optical lattice. The reduced kinetic energy drop upon formation of a polarized phase in an optical lattice extends the ferromagnetic phase to smaller scattering lengths where three-body recombination is small enough to permit experimental detection of the phase. We also show, using time dependent density functional theory, that in such a setup ferromagnetic domains emerge rapidly from a paramagnetic initial state.

cond-mat.quant-gas↗

arXiv2013

The Implications of Diverse Applications and Scalable Data Sets in Benchmarking Big Data Systems

Now we live in an era of big data, and big data applications are becoming more and more pervasive. How to benchmark data center computer systems running big data applications (in short big data systems) is a hot topic. In this paper, we focus on measuring the performance impacts of diverse applications and scalable volumes of data sets on big data systems. For four typical data analysis applications---an important class of big data applications, we find two major results through experiments: first, the data scale has a significant impact on the performance of big data systems, so we must provide scalable volumes of data sets in big data benchmarks. Second, for the four applications, even all of them use the simple algorithms, the performance trends are different with increasing data scales, and hence we must consider not only variety of data sets but also variety of applications in benchmarking big data systems.

cs.PF↗

arXiv2013

Characterizing Data Analysis Workloads in Data Centers

As the amount of data explodes rapidly, more and more corporations are using data centers to make effective decisions and gain a competitive edge. Data analysis applications play a significant role in data centers, and hence it has became increasingly important to understand their behaviors in order to further improve the performance of data center computer systems. In this paper, after investigating three most important application domains in terms of page views and daily visitors, we choose eleven representative data analysis workloads and characterize their micro-architectural characteristics by using hardware performance counters, in order to understand the impacts and implications of data analysis workloads on the systems equipped with modern superscalar out-of-order processors. Our study on the workloads reveals that data analysis applications share many inherent characteristics, which place them in a different class from desktop (SPEC CPU2006), HPC (HPCC), and service workloads, including traditional server workloads (SPECweb2005) and scale-out service workloads (four among six benchmarks in CloudSuite), and accordingly we give several recommendations for architecture and system optimizations. On the basis of our workload characterization work, we released a benchmark suite named DCBench for typical datacenter workloads, including data analysis and service workloads, with an open-source license on our project home page on http://prof.ict.ac.cn/DCBench. We hope that DCBench is helpful for performing architecture and small-to-medium scale system researches for datacenter computing.

cs.PF↗

arXiv2013

Complete Solutions for a Combinatorial Puzzle in Linear Time

In this paper we study a single player game consisting of $n$ black checkers and $m$ white checkers, called shifting the checkers. We have proved that the minimum number of steps needed to play the game for general $n$ and $m$ is $nm + n + m$. We have also presented an optimal algorithm to generate an optimal move sequence of the game consisting of $n$ black checkers and $m$ white checkers, and finally, we present an explicit solution for the general game.

cs.DS↗

arXiv2013

BigDataBench: a Big Data Benchmark Suite from Web Search Engines

This paper presents our joint research efforts on big data benchmarking with several industrial partners. Considering the complexity, diversity, workload churns, and rapid evolution of big data systems, we take an incremental approach in big data benchmarking. For the first step, we pay attention to search engines, which are the most important domain in Internet services in terms of the number of page views and daily visitors. However, search engine service providers treat data, applications, and web access logs as business confidentiality, which prevents us from building benchmarks. To overcome those difficulties, with several industry partners, we widely investigated the open source solutions in search engines, and obtained the permission of using anonymous Web access logs. Moreover, with two years' great efforts, we created a sematic search engine named ProfSearch (available from http://prof.ict.ac.cn). These efforts pave the path for our big data benchmark suite from search engines---BigDataBench, which is released on the web page (http://prof.ict.ac.cn/BigDataBench). We report our detailed analysis of search engine workloads, and present our benchmarking methodology. An innovative data generation methodology and tool are proposed to generate scalable volumes of big data from a small seed of real data, preserving semantics and locality of data. Also, we preliminarily report two case studies using BigDataBench for both system and architecture researches.

cs.IR↗

arXiv2013

How are mortality rates affected by population density?

Biologists have found that the death rate of cells in culture depends upon their spatial density. Permanent "Stay alive" signals from their neighbours seem to prevent them from dying. In a previous paper (Wang et al. 2013) we gave evidence for a density effect for ants. In this paper we examine whether there is a similar effect in human demography. We find that although there is no observable relationship between population density and overall death rates, there is a clear relationship between density and the death rates of young age-groups. Basically their death rates decrease with increasing density. However, this relationship breaks down around 300 inhabitants per square kilometre. Above this threshold the death rates remains fairly constant. The same density effect is observed in Canada, France, Japan and the United States. We also observe a striking parallel between the density effect and the so-called marital status effect in the sense that they both lead to higher suicide rates and are both enhanced for younger age-groups. However, it should be noted that the strength of the density effect is only a fraction of the strength of the marital status effect. In spite of the fact that this parallel does not give us an explanation by itself, it invites us to focus on explanations that apply to both effects. In this light the "Stay alive" paradigm set forth by Prof. Martin Raff appears as a natural interpretation. It can be seen as an extension of the "social ties" framework proposed at the end of the 19th century by the sociologist Emile Durkheim in his study about suicide.

physics.soc-ph↗

arXiv2013

Topological charge pumping in a one-dimensional optical lattice

A topological charge pump [1] transfers charge in a quantized fashion. The quantization is stable against the detailed form of the pumping protocols and external noises and shares the same topological origin as the quantum Hall effect. We propose an experiment setup to realize topological charge pumping of cold atoms in a one-dimensional optical lattice. The quantization of the pumped charge is confirmed by first-principle simulations of the dynamics of uniform and trapped systems. Quantum effects are shown to be crucial for the topological protection of the charge quantization. Finite-temperature and non-adiabatic effect on the experimental observables are discussed. Realization of such a topological charge pump servers as a firm step towards exploring topological states and non-equilibrium dynamics using cold atoms.

cond-mat.quant-gas↗