arXiv subjects

David Brooks

Publications and source records attributed to David Brooks.

At least 217 records · Page 12Linked to original sources

arXiv2020

Candidate Periodically Variable Quasars from the Dark Energy Survey and the Sloan Digital Sky Survey

Periodically variable quasars have been suggested as close binary supermassive black holes. We present a systematic search for periodic light curves in 625 spectroscopically confirmed quasars with a median redshift of 1.8 in a 4.6 deg$^2$ overlapping region of the Dark Energy Survey Supernova (DES-SN) fields and the Sloan Digital Sky Survey Stripe 82 (SDSS-S82). Our sample has a unique 20-year long multi-color ($griz$) light curve enabled by combining DES-SN Y6 observations with archival SDSS-S82 data. The deep imaging allows us to search for periodic light curves in less luminous quasars (down to $r{\sim}$23.5 mag) powered by less massive black holes (with masses $\gtrsim10^{8.5}M_{\odot}$) at high redshift for the first time. We find five candidates with significant (at $>$99.74% single-frequency significance in at least two bands with a global p-value of $\sim$7$\times10^{-4}$--3$\times10^{-3}$ accounting for the look-elsewhere effect) periodicity with observed periods of $\sim$3--5 years (i.e., 1--2 years in rest frame) having $\sim$4--6 cycles spanned by the observations. If all five candidates are periodically variable quasars, this translates into a detection rate of ${\sim}0.8^{+0.5}_{-0.3}$% or ${\sim}1.1^{+0.7}_{-0.5}$ quasar per deg$^2$. Our detection rate is 4--80 times larger than those found by previous searches using shallower surveys over larger areas. This discrepancy is likely caused by differences in the quasar populations probed and the survey data qualities. We discuss implications on the future direct detection of low-frequency gravitational waves. Continued photometric monitoring will further assess the robustness and characteristics of these candidate periodic quasars to determine their physical origins.

astro-ph.HE↗

arXiv2020

Probing galaxy evolution in massive clusters using ACT and DES: splashback as a cosmic clock

We measure the projected number density profiles of galaxies and the splashback feature in clusters selected by the Sunyaev--Zeldovich (SZ) effect from the Advanced Atacama Cosmology Telescope (AdvACT) survey using galaxies observed by the Dark Energy Survey (DES). The splashback radius for the complete galaxy sample is consistent with theoretical measurements from CDM-only simulations, and is located at $2.4^{+0.3}_{-0.4}$ Mpc $h^{-1}$. We split the sample based on galaxy color and find significant differences in the profile shapes. Red galaxies and those in the green valley show a splashback-like minimum in their slope profile consistent with theoretical predictions, while the bluest galaxies show a weak feature that appears at a smaller radius. We develop a mapping of galaxies to subhalos in $N$-body simulations by splitting subhalos based on infall time onto the cluster halos. We find that the location of the steepest slope and differences in the shapes of the profiles can be mapped to differences in the average time of infall of galaxies of different colors. The minima of the slope in the galaxy profiles trace a discontinuity in the phase space of dark matter halos. By relating spatial profiles to infall time for galaxies of different colours, we can use splashback as a clock to understand galaxy quenching. We find that red galaxies have on average been in their clusters for over $3.2 ~\rm Gyrs$, green galaxies about $2.2 ~\rm Gyrs$, while blue galaxies have been accreted most recently and have not reached apocenter. Using the information from the complete radial profiles, we fit a simple quenching model and find that the onset of galaxy quenching in clusters occurs after a delay of about a gigayear, and that galaxies quench rapidly thereafter with an exponential timescale of $0.6$ Gyr.

astro-ph.GA↗

arXiv2020

Imaging Systematics and Clustering of DESI Main Targets

We evaluate the impact of imaging systematics on the clustering of luminous red galaxies (LRG), emission-line galaxies (ELG) and quasars (QSO) targeted for the upcoming Dark Energy Spectroscopic Instrument (DESI) survey. Using Data Release 7 of the DECam Legacy Survey, we study the effects of astrophysical foregrounds, stellar contamination, differences between north galactic cap and south galactic cap measurements, and variations in imaging depth, stellar density, galactic extinction, seeing, airmass, sky brightness, and exposure time before presenting survey masks and weights to mitigate these effects. With our sanitized samples in hand, we conduct a preliminary analysis of the clustering amplitude and evolution of the DESI main targets. From measurements of the angular correlation functions, we determine power law fits $r_0 = 7.78 \pm 0.26$ $h^{-1}$Mpc, $γ= 1.98 \pm 0.02$ for LRGs and $r_0 = 5.45 \pm 0.1$ $h^{-1}$Mpc, $γ= 1.54 \pm 0.01$ for ELGs. Additionally, from the angular power spectra, we measure the linear biases and model the scale dependent biases in the weakly nonlinear regime. Both sets of clustering measurements show good agreement with survey requirements for LRGs and ELGs, attesting that these samples will enable DESI to achieve precise cosmological constraints. We also present clustering as a function of magnitude, use cross-correlations with external spectroscopy to infer $dN/dz$ and measure clustering as a function of luminosity, and probe higher order clustering statistics through counts-in-cells moments.

astro-ph.CO↗

arXiv2020

Dust Reverberation Mapping in Distant Quasars from Optical and Mid-Infrared Imaging Surveys

The size of the dust torus in Active Galactic Nuclei (AGN) and their high-luminosity counterparts, quasars, can be inferred from the time delay between UV/optical accretion disk continuum variability and the response in the mid-infrared (MIR) torus emission. This dust reverberation mapping (RM) technique has been successfully applied to $\sim 70$ $z\lesssim 0.3$ AGN and quasars. Here we present first results of our dust RM program for distant quasars covered in the SDSS Stripe 82 region combining $\sim 20$-yr ground-based optical light curves with 10-yr MIR light curves from the WISE satellite. We measure a high-fidelity lag between W1-band (3.4 $μ$m) and $g$ band for 587 quasars over $0.3\lesssim z\lesssim 2$ ($\left \sim 0.8$) and two orders of magnitude in quasar luminosity. They tightly follow (intrinsic scatter $\sim 0.17$ dex in lag) the IR lag-luminosity relation observed for $z<0.3$ AGN, revealing a remarkable size-luminosity relation for the dust torus over more than four decades in AGN luminosity, with little dependence on additional quasar properties such as Eddington ratio and variability amplitude. This study motivates further investigations in the utility of dust RM for cosmology, and strongly endorses a compelling science case for the combined 10-yr Vera C. Rubin Observatory Legacy Survey of Space and Time (optical) and 5-yr Nancy Grace Roman Space Telescope 2$μ$m light curves in a deep survey for low-redshift AGN dust RM with much lower luminosities and shorter, measurable IR lags. The compiled optical and MIR light curves for 7,384 quasars in our parent sample are made public with this work.

astro-ph.GA↗

arXiv2020

Exploiting Parallelism Opportunities with Deep Learning Frameworks

State-of-the-art machine learning frameworks support a wide variety of design features to enable a flexible machine learning programming interface and to ease the programmability burden on machine learning developers. Identifying and using a performance-optimal setting in feature-rich frameworks, however, involves a non-trivial amount of performance profiling efforts and often relies on domain-specific knowledge. This paper takes a deep dive into analyzing the performance impact of key design features in a machine learning framework and quantifies the role of parallelism. The observations and insights distill into a simple set of guidelines that one can use to achieve much higher training and inference speedup. Across a diverse set of real-world deep learning models, the evaluation results show that the proposed performance tuning guidelines outperform the Intel and TensorFlow recommended settings by 1.29x and 1.34x, respectively.

cs.LG↗

arXiv2020

Dark Energy Survey Identification of A Low-Mass Active Galactic Nucleus at Redshift 0.823 from Optical Variability

We report the identification of a low-mass AGN, DES J0218$-$0430, in a redshift $z = 0.823$ galaxy in the Dark Energy Survey (DES) Supernova field. We select DES J0218$-$0430 as an AGN candidate by characterizing its long-term optical variability alone based on DES optical broad-band light curves spanning over 6 years. An archival optical spectrum from the fourth phase of the Sloan Digital Sky Survey shows both broad Mg II and broad H$β$ lines, confirming its nature as a broad-line AGN. Archival XMM-Newton X-ray observations suggest an intrinsic hard X-ray luminosity of $L_{\rm 2-12\,keV}\sim7.6\pm0.4\times10^{43}$ erg s$^{-1}$, which exceeds those of the most X-ray luminous starburst galaxies, in support of an AGN driving the optical variability. Based on the broad H$β$ from SDSS spectrum, we estimate a virial BH mass of $M_{\bullet}\approx10^{6.43}$-$10^{6.72}M_{\odot}$ (with the error denoting 1$σ$ statistical uncertainties only), consistent with the estimation from OzDES, making it the lowest mass AGN with redshift $>$ 0.4 detected in optical. We estimate the host galaxy stellar mass to be $M_{\ast}\sim10^{10.5\pm0.3}M_{\odot}$ based on modeling the multi-wavelength spectral energy distribution. DES J0218$-$0430 extends the $M_{\bullet}$-$M_{\ast}$ relation observed in luminous AGNs at $z\sim1$ to masses lower than being probed by previous work. Our work demonstrates the feasibility of using optical variability to identify low-mass AGNs at higher redshift in deeper synoptic surveys with direct implications for the upcoming Legacy Survey of Space and Time at Vera C. Rubin Observatory.

astro-ph.GA↗

arXiv2020

CHIPKIT: An agile, reusable open-source framework for rapid test chip development

The current trend for domain-specific architectures (DSAs) has led to renewed interest in research test chips to demonstrate new specialized hardware. Tape-outs also offer huge pedagogical value garnered from real hands-on exposure to the whole system stack. However, successful tape-outs demand hard-earned experience, and the design process is time consuming and fraught with challenges. Therefore, custom chips have remained the preserve of a small number of research groups, typically focused on circuit design research. This paper describes the CHIPKIT framework. We describe a reusable SoC subsystem which provides basic IO, an on-chip programmable host, memory and peripherals. This subsystem can be readily extended with new IP blocks to generate custom test chips. We also present an agile RTL development flow, including a code generation tool calledVGEN. Finally, we outline best practices for full-chip validation across the entire design cycle.

cs.AR↗

arXiv2020

The Curious Case of PHL 293B: A Long-Lived Transient in a Metal-Poor Blue Compact Dwarf Galaxy

We report on small-amplitude optical variability and recent dissipation of the unusually persistent broad emission lines in the blue compact dwarf galaxy PHL 293B. The galaxy's unusual spectral features (P Cygni-like profiles with $\sim$800 km s$^{-1}$ blueshifted absorption lines) have resulted in conflicting interpretations of the nature of this source in the literature. However, analysis of new Gemini spectroscopy reveals the broad emission has begun to fade after being persistent for over a decade prior. Precise difference imaging light curves constructed with the Sloan Digital Sky Survey and the Dark Energy Survey reveal small-amplitude optical variability of $\sim$0.1 mag in the g band offset by $100\pm21$ pc from the brightest pixel of the host. The light curve is well-described by an active galactic nuclei (AGN)-like damped random walk process. However, we conclude that the origin of the optical variability and spectral features of PHL 293B is due to a long-lived stellar transient, likely a Type IIn supernova or non-terminal outburst, mimicking long-term AGN-like variability. This work highlights the challenges of discriminating between scenarios in such extreme environments, relevant to searches for AGNs in dwarf galaxies. This is the second long-lived transient discovered in a blue compact dwarf, after SDSS1133. Our result implies such long-lived stellar transients may be more common in metal-deficient galaxies. Systematic searches for low-level variability in dwarf galaxies will be possible with the upcoming Legacy Survey of Space and Time at Vera C. Rubin Observatory.

astro-ph.GA↗

arXiv2020

MLPerf Training Benchmark

Machine learning (ML) needs industry-standard performance benchmarks to support design and competitive evaluation of the many emerging software and hardware solutions for ML. But ML training presents three unique benchmarking challenges absent from other domains: optimizations that improve training throughput can increase the time to solution, training is stochastic and time to solution exhibits high variance, and software and hardware systems are so diverse that fair benchmarking with the same binary, code, and even hyperparameters is difficult. We therefore present MLPerf, an ML benchmark that overcomes these challenges. Our analysis quantitatively evaluates MLPerf's efficacy at driving performance and scalability improvements across two rounds of results from multiple vendors.

cs.LG↗

arXiv2020

Dark Energy Survey Year 1 Results: Cosmological Constraints from Cluster Abundances and Weak Lensing

We perform a joint analysis of the counts and weak lensing signal of redMaPPer clusters selected from the Dark Energy Survey (DES) Year 1 dataset. Our analysis uses the same shear and source photometric redshifts estimates as were used in the DES combined probes analysis. Our analysis results in surprisingly low values for $S_8 =σ_8(Ω_{\rm m}/0.3)^{0.5}= 0.65\pm 0.04$, driven by a low matter density parameter, $Ω_{\rm m}=0.179^{+0.031}_{-0.038}$, with $σ_8-Ω_{\rm m}$ posteriors in $2.4σ$ tension with the DES Y1 3x2pt results, and in $5.6σ$ with the Planck CMB analysis. These results include the impact of post-unblinding changes to the analysis, which did not improve the level of consistency with other data sets compared to the results obtained at the unblinding. The fact that multiple cosmological probes (supernovae, baryon acoustic oscillations, cosmic shear, galaxy clustering and CMB anisotropies), and other galaxy cluster analyses all favor significantly higher matter densities suggests the presence of systematic errors in the data or an incomplete modeling of the relevant physics. Cross checks with X-ray and microwave data, as well as independent constraints on the observable--mass relation from SZ selected clusters, suggest that the discrepancy resides in our modeling of the weak lensing signal rather than the cluster abundance. Repeating our analysis using a higher richness threshold ($λ\ge 30$) significantly reduces the tension with other probes, and points to one or more richness-dependent effects not captured by our model.

astro-ph.CO↗

arXiv2020

Optimising Automatic Morphological Classification of Galaxies with Machine Learning and Deep Learning using Dark Energy Survey Imaging

There are several supervised machine learning methods used for the application of automated morphological classification of galaxies; however, there has not yet been a clear comparison of these different methods using imaging data, or a investigation for maximising their effectiveness. We carry out a comparison between several common machine learning methods for galaxy classification (Convolutional Neural Network (CNN), K-nearest neighbour, Logistic Regression, Support Vector Machine, Random Forest, and Neural Networks) by using Dark Energy Survey (DES) data combined with visual classifications from the Galaxy Zoo 1 project (GZ1). Our goal is to determine the optimal machine learning methods when using imaging data for galaxy classification. We show that CNN is the most successful method of these ten methods in our study. Using a sample of $\sim$2,800 galaxies with visual classification from GZ1, we reach an accuracy of $\sim$0.99 for the morphological classification of Ellipticals and Spirals. The further investigation of the galaxies that have a different ML and visual classification but with high predicted probabilities in our CNN usually reveals an the incorrect classification provided by GZ1. We further find the galaxies having a low probability of being either spirals or ellipticals are visually Lenticulars (S0), demonstrating that supervised learning is able to rediscover that this class of galaxy is distinct from both Es and Spirals. We confirm that $\sim$2.5\% galaxies are misclassified by GZ1 in our study. After correcting these galaxies' labels, we improve our CNN performance to an average accuracy of over 0.99 (accuracy of 0.994 is our best result).

astro-ph.GA↗

arXiv2020

The Architectural Implications of Facebook's DNN-based Personalized Recommendation

The widespread application of deep learning has changed the landscape of computation in the data center. In particular, personalized recommendation for content ranking is now largely accomplished leveraging deep neural networks. However, despite the importance of these models and the amount of compute cycles they consume, relatively little research attention has been devoted to systems for recommendation. To facilitate research and to advance the understanding of these workloads, this paper presents a set of real-world, production-scale DNNs for personalized recommendation coupled with relevant performance metrics for evaluation. In addition to releasing a set of open-source workloads, we conduct in-depth analysis that underpins future system design and optimization for at-scale recommendation: Inference latency varies by 60% across three Intel server generations, batching and co-location of inferences can drastically improve latency-bounded throughput, and the diverse composition of recommendation models leads to different optimization strategies.

cs.DC↗

arXiv2020

AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference

Conventional hardware-friendly quantization methods, such as fixed-point or integer, tend to perform poorly at very low word sizes as their shrinking dynamic ranges cannot adequately capture the wide data distributions commonly seen in sequence transduction models. We present AdaptivFloat, a floating-point inspired number representation format for deep learning that dynamically maximizes and optimally clips its available dynamic range, at a layer granularity, in order to create faithful encoding of neural network parameters. AdaptivFloat consistently produces higher inference accuracies compared to block floating-point, uniform, IEEE-like float or posit encodings at very low precision ($\leq$ 8-bit) across a diverse set of state-of-the-art neural network topologies. And notably, AdaptivFloat is seen surpassing baseline FP32 performance by up to +0.3 in BLEU score and -0.75 in word error rate at weight bit widths that are $\leq$ 8-bit. Experimental results on a deep neural network (DNN) hardware accelerator, exploiting AdaptivFloat logic in its computational datapath, demonstrate per-operation energy and area that is 0.9$\times$ and 1.14$\times$, respectively, that of equivalent bit width integer-based accelerator variants.

cs.LG↗

arXiv2020

DeepRecSys: A System for Optimizing End-To-End At-scale Neural Recommendation Inference

Neural personalized recommendation is the corner-stone of a wide collection of cloud services and products, constituting significant compute demand of the cloud infrastructure. Thus, improving the execution efficiency of neural recommendation directly translates into infrastructure capacity saving. In this paper, we devise a novel end-to-end modeling infrastructure, DeepRecInfra, that adopts an algorithm and system co-design methodology to custom-design systems for recommendation use cases. Leveraging the insights from the recommendation characterization, a new dynamic scheduler, DeepRecSched, is proposed to maximize latency-bounded throughput by taking into account characteristics of inference query size and arrival patterns, recommendation model architectures, and underlying hardware systems. By doing so, system throughput is doubled across the eight industry-representative recommendation models. Finally, design, deployment, and evaluation in at-scale production datacenter shows over 30% latency reduction across a wide variety of recommendation models running on hundreds of machines.

cs.DC↗

arXiv2019

RecNMP: Accelerating Personalized Recommendation with Near-Memory Processing

Personalized recommendation systems leverage deep learning models and account for the majority of data center AI cycles. Their performance is dominated by memory-bound sparse embedding operations with unique irregular memory access patterns that pose a fundamental challenge to accelerate. This paper proposes a lightweight, commodity DRAM compliant, near-memory processing solution to accelerate personalized recommendation inference. The in-depth characterization of production-grade recommendation models shows that embedding operations with high model-, operator- and data-level parallelism lead to memory bandwidth saturation, limiting recommendation inference performance. We propose RecNMP which provides a scalable solution to improve system throughput, supporting a broad range of sparse embedding models. RecNMP is specifically tailored to production environments with heavy co-location of operators on a single server. Several hardware/software co-optimization techniques such as memory-side caching, table-aware packet scheduling, and hot entry profiling are studied, resulting in up to 9.8x memory latency speedup over a highly-optimized baseline. Overall, RecNMP offers 4.2x throughput improvement and 45.8% memory energy savings.

cs.DC↗

arXiv2019

SMAUG: End-to-End Full-Stack Simulation Infrastructure for Deep Learning Workloads

In recent years, there has been tremendous advances in hardware acceleration of deep neural networks. However, most of the research has focused on optimizing accelerator microarchitecture for higher performance and energy efficiency on a per-layer basis. We find that for overall single-batch inference latency, the accelerator may only make up 25-40%, with the rest spent on data movement and in the deep learning software framework. Thus far, it has been very difficult to study end-to-end DNN performance during early stage design (before RTL is available) because there are no existing DNN frameworks that support end-to-end simulation with easy custom hardware accelerator integration. To address this gap in research infrastructure, we present SMAUG, the first DNN framework that is purpose-built for simulation of end-to-end deep learning applications. SMAUG offers researchers a wide range of capabilities for evaluating DNN workloads, from diverse network topologies to easy accelerator modeling and SoC integration. To demonstrate the power and value of SMAUG, we present case studies that show how we can optimize overall performance and energy efficiency for up to 1.8-5x speedup over a baseline system, without changing any part of the accelerator microarchitecture, as well as show how SMAUG can tune an SoC for a camera-powered deep learning pipeline.

cs.LG↗

arXiv2019

Benchmarking TPU, GPU, and CPU Platforms for Deep Learning

Training deep learning models is compute-intensive and there is an industry-wide trend towards hardware specialization to improve performance. To systematically benchmark deep learning platforms, we introduce ParaDnn, a parameterized benchmark suite for deep learning that generates end-to-end models for fully connected (FC), convolutional (CNN), and recurrent (RNN) neural networks. Along with six real-world models, we benchmark Google's Cloud TPU v2/v3, NVIDIA's V100 GPU, and an Intel Skylake CPU platform. We take a deep dive into TPU architecture, reveal its bottlenecks, and highlight valuable lessons learned for future specialized system design. We also provide a thorough comparison of the platforms and find that each has unique strengths for some types of models. Finally, we quantify the rapid performance improvements that specialized software stacks provide for the TPU and GPU platforms.

cs.LG↗

arXiv2019

MASR: A Modular Accelerator for Sparse RNNs

Recurrent neural networks (RNNs) are becoming the de facto solution for speech recognition. RNNs exploit long-term temporal relationships in data by applying repeated, learned transformations. Unlike fully-connected (FC) layers with single vector matrix operations, RNN layers consist of hundreds of such operations chained over time. This poses challenges unique to RNNs that are not found in convolutional neural networks (CNNs) or FC models, namely large dynamic activation. In this paper we present MASR, a principled and modular architecture that accelerates bidirectional RNNs for on-chip ASR. MASR is designed to exploit sparsity in both dynamic activations and static weights. The architecture is enhanced by a series of dynamic activation optimizations that enable compact storage, ensure no energy is wasted computing null operations, and maintain high MAC utilization for highly parallel accelerator designs. In comparison to current state-of-the-art sparse neural network accelerators (e.g., EIE), MASR provides 2x area 3x energy, and 1.6x performance benefits. The modular nature of MASR enables designs that efficiently scale from resource-constrained low-power IoT applications to large-scale, highly parallel datacenter deployments.

eess.SP↗