Showing papers by "IBM published in 2010"

Adaptive Subgradient Methods for Online Learning and Stochastic Optimization.

Abstract: We present a new family of subgradient methods that dynamically incorporate knowledge of the geometry of the data observed in earlier iterations to perform more informative gradient-based learning. Metaphorically, the adaptation allows us to find needles in haystacks in the form of very predictive but rarely seen features. Our paradigm stems from recent advances in stochastic optimization and online learning which employ proximal functions to control the gradient steps of the algorithm. We describe and analyze an apparatus for adaptively modifying the proximal function, which significantly simplifies setting a learning rate and results in regret guarantees that are provably as good as the best proximal function that can be chosen in hindsight. We give several efficient algorithms for empirical risk minimization problems with common and important regularization functions and domain constraints. We experimentally study our theoretical analysis and show that adaptive subgradient methods outperform state-of-the-art, yet non-adaptive, subgradient algorithms.

An Introduction to Symbolic Dynamics and Coding

Abstract: From the Publisher: Although it originated as a method to study general dynamical systems, symbolic dynamics is useful in coding for data storage and transmission as well as in linear algebra. Requiring only a undergraduate knowledge of linear algebra, this first general textbook includes over 500 exercises.

100-GHz Transistors from Wafer-Scale Epitaxial Graphene

Abstract: The high carrier mobility of graphene has been exploited in field-effect transistors that operate at high frequencies. Transistors were fabricated on epitaxial graphene synthesized on the silicon face of a silicon carbide wafer, achieving a cutoff frequency of 100 gigahertz for a gate length of 240 nanometers. The high-frequency performance of these epitaxial graphene transistors exceeds that of state-of-the-art silicon transistors of the same gate length.

Lithium−Air Battery: Promise and Challenges

Abstract: The lithium−air system captured worldwide attention in 2009 as a possible battery for electric vehicle propulsion applications. If successfully developed, this battery could provide an energy source for electric vehicles rivaling that of gasoline in terms of usable energy density. However, there are numerous scientific and technical challenges that must be overcome if this alluring promise is to turn into reality. The fundamental battery chemistry during discharge is thought to be the electrochemical oxidation of lithium metal at the anode and reduction of oxygen from air at the cathode. With aprotic electrolytes, as used in Li-ion batteries, there is some evidence that the process can be reversed by applying an external potential, i.e., that such a battery can be electrically recharged. This paper summarizes the authors’ view of the promise and challenges facing development of practical Li−air batteries and the current understanding of its chemistry. However, it must be appreciated that this perspective ...

Graphene photodetectors for high-speed optical communications

Abstract: Although silicon has dominated solid-state electronics for more than four decades, a variety of other materials are used in photonic devices to expand the wavelength range of operation and improve performance. For example, gallium-nitride based materials enable light emission at blue and ultraviolet wavelengths1, and high index contrast silicon-on-insulator facilitates ultradense photonic devices2,3. Here, we report the first use of a photodetector based on graphene4,5, a two-dimensional carbon material, in a 10 Gbit s−1 optical data link. In this interdigitated metal–graphene–metal photodetector, an asymmetric metallization scheme is adopted to break the mirror symmetry of the internal electric-field profile in conventional graphene field-effect transistor channels6,7,8,9, allowing for efficient photodetection. A maximum external photoresponsivity of 6.1 mA W−1 is achieved at a wavelength of 1.55 µm. Owing to the unique band structure of graphene10,11 and extensive developments in graphene electronics12,13 and wafer-scale synthesis13, graphene-based integrated electronic–photonic circuits with an operational wavelength range spanning 300 nm to 6 µm (and possibly beyond) can be expected in the future. A graphene-based photodetector with unprecedented photoresponsivity and the ability to perform error-free detection of 10 Gbit s−1s data streams is demonstrated. The results suggest that graphene-based photonic devices have a bright future in telecommunications and other optical applications.

Phase Change Memory

Abstract: In this paper, recent progress of phase change memory (PCM) is reviewed. The electrical and thermal properties of phase change materials are surveyed with a focus on the scalability of the materials and their impact on device design. Innovations in the device structure, memory cell selector, and strategies for achieving multibit operation and 3-D, multilayer high-density memory arrays are described. The scaling properties of PCM are illustrated with recent experimental results using special device test structures and novel material synthesis. Factors affecting the reliability of PCM are discussed.

Fully homomorphic encryption over the integers

Abstract: We construct a simple fully homomorphic encryption scheme, using only elementary modular arithmetic. We use Gentry’s technique to construct a fully homomorphic scheme from a “bootstrappable” somewhat homomorphic scheme. However, instead of using ideal lattices over a polynomial ring, our bootstrappable encryption scheme merely uses addition and multiplication over the integers. The main appeal of our scheme is the conceptual simplicity. We reduce the security of our scheme to finding an approximate integer gcd – i.e., given a list of integers that are near-multiples of a hidden integer, output that hidden integer. We investigate the hardness of this task, building on earlier work of Howgrave-Graham.

Building Watson: An Overview of the DeepQA Project

Abstract: IBM Research undertook a challenge to build a computer system that could compete at the human champion level in real time on the American TV Quiz show, Jeopardy! The extent of the challenge includes fielding a real-time automatic contestant on the show, not merely a laboratory exercise. The Jeopardy! Challenge helped us address requirements that led to the design of the DeepQA architecture and the implementation of Watson. After 3 years of intense research and development by a core team of about 20 researches, Watson is performing at human expert-levels in terms of precision, confidence and speed at the Jeopardy! Quiz show. Our results strongly suggest that DeepQA is an effective and extensible architecture that may be used as a foundation for combining, deploying, evaluating and advancing a wide range of algorithmic techniques to rapidly advance the field of QA.

Graphene: electronic and photonic properties and devices.

Abstract: Graphene is in many respects a nanomaterial with unique properties. Here I discuss the electronic structure, transport and optical properties of graphene, and how these are utilized in exploratory electronic and optoelectronic devices. Some suggestions for needed advances are made.

Graphene field-effect transistors with high on/off current ratio and large transport band gap at room temperature.

Abstract: Graphene is considered to be a promising candidate for future nanoelectronics due to its exceptional electronic properties. Unfortunately, the graphene field-effect transistors (FETs) cannot be turned off effectively due to the absence of a band gap, leading to an on/off current ratio typically around 5 in top-gated graphene FETs. On the other hand, theoretical investigations and optical measurements suggest that a band gap up to a few hundred millielectronvolts can be created by the perpendicular E-field in bilayer graphenes. Although previous carrier transport measurements in bilayer graphene transistors did indicate a gate-induced insulating state at temperatures below 1 K, the electrical (or transport) band gap was estimated to be a few millielectronvolts, and the room temperature on/off current ratio in bilayer graphene FETs remains similar to those in single-layer graphene FETs. Here, for the first time, we report an on/off current ratio of around 100 and 2000 at room temperature and 20 K, respectively, in our dual-gate bilayer graphene FETs. We also measured an electrical band gap of >130 and 80 meV at average electric displacements of 2.2 and 1.3 V nm(-1), respectively. This demonstration reveals the great potential of bilayer graphene in applications such as digital electronics, pseudospintronics, terahertz technology, and infrared nanophotonics.

High‐Efficiency Solar Cell with Earth‐Abundant Liquid‐Processed Absorber

Abstract: 2010 WILEY-VCH Verlag Gmb Chalcogenide-based solar cells provide a critical pathway to cost parity between photovoltaic (PV) and conventional energy sources. Currently, only Cu(In,Ga)(S,Se)2 (CIGS) and CdTe technologies have reached commercial module production with stable power conversion efficiencies of over 9 percent. Despite the promise of these technologies, restrictions on heavy metal usage for Cd and limitations in supply for In and Te are projected to restrict the production capacity of the existing chalcogen-based technologies to <100GWp per year, a small fraction of our growing energy needs, which are expected to double to 27 TW by 2050. Earth-abundant copper-zinc-tin-chalcogenide kesterites, Cu2ZnSnS4 and Cu2ZnSnSe4, have been examined as potential alternatives for the two leading technologies, reaching promising but not yet marketable efficiencies of 6.7% and 3.2%, respectively, by multilayer vacuum deposition. Here we show a non-vacuum, slurry-based coating method that combines advantages of both solution processing and particlebased deposition, enabling fabrication of Cu2ZnSn(Se,S)4 devices with over 9.6% efficiency—a factor of five performance improvement relative to previous attempts to use highthroughput ink-based approaches and >40% higher than previous record devices prepared using vacuum-based methods. To address the issue of cost, non-vacuum ‘‘ink’’-based approaches—both from solutions and suspensions—are being developed for chalcogenide-based absorber layer deposition to replace potentially more expensive vacuum-based techniques. True solutions allow intermixing of the constituents at a molecular level and the formation of smooth homogeneous films, as demonstrated with spin-coated CIGS absorber layers from hydrazine (N2H4) solutions. [11–13] The chemically reducing character of hydrazine stabilizes solutions of anions with direct metal-chalcogen bonding for select elements (e.g. Cu, In, Ga, Sn), without the necessity to introduce typical impurities (e.g., C, O, Cl). Suspension approaches employ solid particles designed to be deposited on a substrate and reacted or fused with each other, to form a desired crystalline phase and grain structure. Normally insoluble components can be deposited by this approach using typical liquid-based deposition (e.g., printing, spin coating, slit casting, spraying). Although high-quality large-grained absorber layers can be formed for selected systems using either solutionor particlebased deposition, numerous challenges confront each approach for more general deposition needs. Solution processing is limited by the solubility of many materials of interest (e.g., ZnSe1–xSx in hydrazine solvents—relevant for the deposition of Cu2ZnSnS4 or Cu2ZnSnSe4). In addition, volume contraction upon drying of solution-deposited layers creates stress in the film that may cause crack formation in thicker films. In suspension approaches, a common difficulty is achieving single-phase crystallization among the solid particles. Particle-based approaches (as well as some solution methods) typically require the addition of organic agents to improve wetting and particle dispersion, and to avoid film cracks and delamination. Most of these non-volatile organic additives introduce carbon contamination in the final layer. Because of these challenges, vacuum-based techniques have historically shown superior performance to liquid coating. In the case of the earth-abundant Cu2ZnSn(S,Se)4 materials, ink-based approaches have to date yielded at most <1.6% efficiency devices. Here we demonstrate an hybrid solution-particle approach, using the earth-abundant Cu2ZnSn(S,Se)4 system as an example, which enables fabrication of PV devices with over 9.6% power conversion efficiency. The slurry (or ink) employed for deposition comprises a Cu–Sn chalcogenide (S or S–Se) solution in hydrazine (see Experimental section), with the in situ formation of readily dispersible particle-based Zn-chalcogenide precursors, ZnSe(N2H4) (Figure 1a,d) or ZnS(N2H4) (Figure 1b). Thermogravimetric analysis (TGA) of the isolated selenide particle precursor shows decomposition at approximately 200 8C, with mass loss of about 20%, close to the theoretical value expected upon transition to pure ZnSe (Figure 1c,d). Deposition using this hybrid slurry successfully combines the advantages of solution and suspension deposition routes by use of solutions containing solid particles, wherein both components (i.e., solution and particle) contain metal and chalcogen elements that integrate into the final film. Using the hybrid slurry method (i) solubility limitations are resolved, as virtually any materials system can be constituted by a combination of solid and dissolved components; (ii) the dissolved components can be engineered as an efficient binding media for the particles, eliminating the need of separate organic binders; (iii) solid particles act as stress-relief and crack-deflection centers allowing the deposition of thicker layers than pure solution processes; and (iv) the intimate contact between the two phases allows rapid reaction and homogeneous phase formation. Complete conversion of all constituents of the spin-coated hybrid precursor films into a single-phase, highly crystalline Cu2ZnSn(S,Se)4 is achieved by annealing at 540 8C on a hot plate. Three main types of samples were targeted – high selenium content (A), intermediate sulfoselenide (B) and pure sulfide (C) –

Non-interactive verifiable computing: outsourcing computation to untrusted workers

Abstract: We introduce and formalize the notion of Verifiable Computation, which enables a computationally weak client to "outsource" the computation of a function F on various dynamically-chosen inputs x1, ...,xk to one or more workers. The workers return the result of the function evaluation, e.g., yi = F(xi), as well as a proof that the computation of F was carried out correctly on the given value xi. The primary constraint is that the verification of the proof should require substantially less computational effort than computing F(i) from scratch. We present a protocol that allows the worker to return a computationally-sound, non-interactive proof that can be verified in O(mċpoly(λ)) time, where m is the bit-length of the output of F, and λ is a security parameter. The protocol requires a one-time pre-processing stage by the client which takes O(|C|ċpoly(λ)) time, where C is the smallest known Boolean circuit computing F. Unlike previous work in this area, our scheme also provides (at no additional cost) input and output privacy for the client, meaning that the workers do not learn any information about the xi or yi values.

Improving the Scalability of Data Center Networks with Traffic-aware Virtual Machine Placement

Abstract: The scalability of modern data centers has become a practical concern and has attracted significant attention in recent years. In contrast to existing solutions that require changes in the network architecture and the routing protocols, this paper proposes using traffic-aware virtual machine (VM) placement to improve the network scalability. By optimizing the placement of VMs on host machines, traffic patterns among VMs can be better aligned with the communication distance between them, e.g. VMs with large mutual bandwidth usage are assigned to host machines in close proximity. We formulate the VM placement as an optimization problem and prove its hardness. We design a two-tier approximate algorithm that efficiently solves the VM placement problem for very large problem sizes. Given the significant difference in the traffic patterns seen in current data centers and the structural differences of the recently proposed data center architectures, we further conduct a comparative analysis on the impact of the traffic patterns and the network architectures on the potential performance gain of traffic-aware VM placement. We use traffic traces collected from production data centers to evaluate our proposed VM placement algorithm, and we show a significant performance improvement compared to existing general methods that do not take advantage of traffic patterns and data center network characteristics.

Foundations for smarter cities

Abstract: This paper describes the information technology (IT) foundation and principles for Smarter Cities™. Smarter Cities are urban areas that exploit operational data, such as that arising from traffic congestion, power consumption statistics, and public safety events, to optimize the operation of city services. The foundational concepts are instrumented, interconnected, and intelligent. Instrumented refers to sources of near-real-time real-world data from both physical and virtual sensors. Interconnected means the integration of those data into an enterprise computing platform and the communication of such information among the various city services. Intelligent refers to the inclusion of complex analytics, modeling, optimization, and visualization in the operational business processes to make better operational decisions. This approach enables the adaptation of city services to the behavior of the inhabitants, which permits the optimal use of the available physical infrastructure and resources, for example, in sensing and controlling consumption of energy and water, managing waste processing and transportation systems, and applying optimization to achieve new efficiencies among these resources. Additional roles exist in intelligent interaction between the city and its inhabitants and further contribute to operational efficiency while maintaining or enhancing quality of life.

Phase change memory technology

Abstract: The authors survey the current state of phase change memory (PCM), a nonvolatile solid-state memory technology built around the large electrical contrast between the highly resistive amorphous and highly conductive crystalline states in so-called phase change materials. PCM technology has made rapid progress in a short time, having passed older technologies in terms of both sophisticated demonstrations of scaling to small device dimensions, as well as integrated large-array demonstrators with impressive retention, endurance, performance, and yield characteristics. They introduce the physics behind PCM technology, assess how its characteristics match up with various potential applications across the memory-storage hierarchy, and discuss its strengths including scalability and rapid switching speed. Challenges for the technology are addressed, including the design of PCM cells for low reset current, the need to control device-to-device variability, and undesirable changes in the phase change material that c...

Block copolymer based nanostructures: materials, processes, and applications to electronics.

Abstract: 2.4. Block Copolymer Containing Hybrids 151 3. Block Copolymer Ordering in Thin Films 153 3.1. General Process Steps 153 3.2. Morphology of Thin Films 154 3.3. Thickness-Dependent Nanopatterning 154 3.3.1. Ultrathin Films: Monomolecular Films 155 3.3.2. Sub-L0 Thick Films 155 3.3.3. Thick Films 157 3.4. Placement Control: Directed Self-Assembly (DSA) 162 3.4.1. Topographic Guiding Patterns: Graphoepitaxy 163

Revealing strengths and weaknesses of methods for gene network inference

Abstract: Numerous methods have been developed for inferring gene regulatory networks from expression data, however, both their absolute and comparative performance remain poorly understood. In this paper, we introduce a framework for critical performance assessment of methods for gene network inference. We present an in silico benchmark suite that we provided as a blinded, community-wide challenge within the context of the DREAM (Dialogue on Reverse Engineering Assessment and Methods) project. We assess the performance of 29 gene-network-inference methods, which have been applied independently by participating teams. Performance profiling reveals that current inference methods are affected, to various degrees, by different types of systematic prediction errors. In particular, all but the best-performing method failed to accurately infer multiple regulatory inputs (combinatorial regulation) of genes. The results of this community-wide experiment show that reliable network inference from gene expression data remains an unsolved problem, and they indicate potential ways of network reconstruction improvements.

Organocatalysis: Opportunities and Challenges for Polymer Synthesis

Abstract: Organocatalysis offers a number of opportunities in polymer synthesis and was among the earliest methods of catalyzing the synthesis of polyesters. In the following Perspective we attempt to highlight the opportunities and challenges in the use of organic molecules as catalysts or initiators for polymerization reactions. The ring-opening polymerization of cyclic monomers is used as a representative polymerization process to illustrate some of the features of organic catalysts and initiators and to compare them to metal-based approaches. The convergence of convenience, functional group tolerance, fast rates, and selectivities will continue to drive innovations in polymerization catalysis, and it is our perspective that organocatalysis will continue to play an important role in these developments.

Low dielectric constant materials.

Abstract: Modern computer microprocessor chips are marvels of engineering complexity. For the current 45 nm technology node, there may be nearly a billion transistors on a chip barely 1 cm2 and more than 10 000 m of wiring connecting and powering these devices distributed over 9-10 wiring levels. This represents quite an advance from the first INTEL 4004B microprocessor chip introduced in 1971 with 10 μm minimum dimensions and 2 300 transistors on the chip! It has been disclosed that advanced microprocessor chips at the 32 nm node will have more than 2 billion transistors.1 For instance, Figure 1 shows a sectional 3D image of a 90 nm IBM microprocessor, containing several hundred million integrated devices and 10 levels of interconnect wiring, designated as the back-end-of-the-line (BEOL). Since the invention of microprocessors, the number of active devices on a chip has been exponentially increasing, approximately doubling every two years. This trend was first described in 1965 by Gordon Moore,2 although the original discussion suggested doubling the number of devices every year, and the phenomenon became popularly known as Moore’s Law. This progress has proven remarkably resilient and has persisted for more than 50 years. The enabler that has permitted these advances is known as scaling, that is, the reduction of minimum device dimensions by lithographic advances (photoresists, tooling, and process integration optimization) by ∼30% for each device generation.3 It allowed more active devices to be incorporated in a given area and improved the operating characteristics of the individual transistors. It should be emphasized that the earlier improvements in chip performance were achieved with very few changes in the materials used in the construction of the chips themselves. The increase of performance with scaling * Corresponding author. E-mail: gdubois@us.ibm.com. † IBM Almaden Research Center. ‡ Stanford University. Willi Volksen received his B.S. in Chemistry (magna cum laude) from New Mexico Institute of Mining and Technology in 1972 and his Ph.D. in Chemistry/Polymer Science from the University of Massachusetts, Lowell, in 1975. He then joined the research group of Prof. Harry Gray/Dr. Alan Rembaum at the California Institute of Technology as a postdoctoral fellow and upon completion of the one-year appointment joined Dr. Rembaum at the Jet Propulsion Laboratory as a Senior Chemist in 1976. In 1977 Dr. Volksen joined the IBM Research Division at the IBM Almaden Research Center in San Jose, CA, where he is an active research staff member in the Advanced Materials Group of the Science and Technology function.

Phase Change Materials and Their Application to Nonvolatile Memories

Abstract: Phase change materials are materials that exist in at least two structurally distinct solid phases, an amorphous and one (or more) crystalline phases. Many materials display phase change properties in this sense and can be deposited at least as a thin film in an amorphous phase (low temperature deposition, very thin film) or crystalline phase (high temperature deposition, epitaxy). Often the amorphous and crystalline phases have very different optical and electrical properties stemming from the large differences in structure between the amorphous and the crystalline phases. These differences can be used to store information in technological applications if it is possible to switch the material repeatedly between the two phases and if both phases are stable at operating temperature. The transformation of the metastable amorphous phase to the energetically favorable, stable crystalline phase occurs by heating the material above its crystallization temperature for a time

Computing arbitrary functions of encrypted data

Abstract: Suppose that you want to delegate the ability to process your data, without giving away access to it. We show that this separation is possible: we describe a "fully homomorphic" encryption scheme that keeps data private, but that allows a worker that does not have the secret decryption key to compute any (still encrypted) result of the data, even when the function of the data is very complex. In short, a third party can perform complicated processing of data without being able to see it. Among other things, this helps make cloud computing compatible with privacy.

An empirical comparison of machine learning models for time series forecasting

Abstract: In this work we present a large scale comparison study for the major machine learning models for time series forecasting. Specifically, we apply the models on the monthly M3 time series competition data (around a thousand time series). There have been very few, if any, large scale comparison studies for machine learning models for the regression or the time series forecasting problems, so we hope this study would fill this gap. The models considered are multilayer perceptron, Bayesian neural networks, radial basis functions, generalized regression neural networks (also called kernel regression), K-nearest neighbor regression, CART regression trees, support vector regression, and Gaussian processes. The study reveals significant differences between the different methods. The best two methods turned out to be the multilayer perceptron and the Gaussian process regression. In addition to model comparisons, we have tested different preprocessing methods and have shown that they have different impacts on the pe...

Reinventing germanium avalanche photodetector for nanophotonic on-chip optical interconnects

Abstract: A key element in the integration of microprocessor chips with optical communications circuits is a photodetector to mediate the optical and electronic signals. Germanium photodetectors are very attractive in this regard because they are compatible with conventional silicon circuitry, but they suffer from noise that limits their performance. Assefa et al. now show how the poor intrinsic noise characteristics of germanium can be overcome through the careful engineering of optical and electrical fields at the nanometre scale. The result is a compact and efficient photodetector that could enable a range of optoelectronic applications. To integrate microchips with optical communications a photodetector is required to mediate the optical and electronic signals. Although germanium photodetectors are compatible with silicon their performance is impaired by poor intrinsic noise. Here the noise is reduced by nanometre engineering of optical and electrical fields to produce a compact and efficient photodetector. Integration of optical communication circuits directly into high-performance microprocessor chips can enable extremely powerful computer systems1. A germanium photodetector that can be monolithically integrated with silicon transistor technology2,3,4,5,6,7,8 is viewed as a key element in connecting chip components with infrared optical signals. Such a device should have the capability to detect very-low-power optical signals at very high speed. Although germanium avalanche photodetectors9,10 (APD) using charge amplification close to avalanche breakdown can achieve high gain and thus detect low-power optical signals, they are universally considered to suffer from an intolerably high amplification noise characteristic of germanium11. High gain with low excess noise has been demonstrated using a germanium layer only for detection of light signals, with amplification taking place in a separate silicon layer12. However, the relatively thick semiconductor layers that are required in such structures limit APD speeds to about 10 GHz, and require excessively high bias voltages of around 25 V (ref. 12). Here we show how nanophotonic and nanoelectronic engineering aimed at shaping optical and electrical fields on the nanometre scale within a germanium amplification layer can overcome the otherwise intrinsically poor noise characteristics, achieving a dramatic reduction of amplification noise by over 70 per cent. By generating strongly non-uniform electric fields, the region of impact ionization in germanium is reduced to just 30 nm, allowing the device to benefit from the noise reduction effects13,14,15 that arise at these small distances. Furthermore, the smallness of the APDs means that a bias voltage of only 1.5 V is required to achieve an avalanche gain of over 10 dB with operational speeds exceeding 30 GHz. Monolithic integration of such a device into computer chips might enable applications beyond computer optical interconnects1—in telecommunications16, secure quantum key distribution17, and subthreshold ultralow-power transistors18.

The spitzer survey of stellar structure in galaxies (s^4g)

Abstract: The Spitzer Survey of Stellar Structure in Galaxies (S^4G) is an Exploration Science Legacy Program approved for the Spitzer post–cryogenic mission. It is a volume-, magnitude-, and size-limited (d < 40 Mpc, |b|> 30°,m_(Bcorr) 1') survey of 2331 galaxies using the Infrared Array Camera (IRAC) at 3.6 and 4.5 μm. Each galaxy is observed for 240 s and mapped to ≥ 1:5 × D_(25). The final mosaicked images have a typical 1 σ rms noise level of 0.0072 and 0:0093 MJy sr^-1 at 3.6 and 4.5 μm, respectively. Our azimuthally averaged surface brightness profile typically traces isophotes at μ_(3.6μm (AB)(1σ) ~ 27 mag arcsec^(-2), equivalent to a stellar mass surface density of ~1 M_⊙pc^(-2). S^4G thus provides an unprecedented data set for the study of the distribution of mass and stellar structures in the local universe. This large, unbiased, and extremely deep sample of all Hubble types from dwarfs to spirals to ellipticals will allow for detailed structural studies, not only as a function of stellar mass, but also as a function of the local environment. The data from this survey will serve as a vital testbed for cosmological simulations predicting the stellar mass properties of present-day galaxies. This article introduces the survey and describes the sample selection, the significance of the 3.6 and 4.5 μm bands for this study, and the data collection and survey strategies. We describe the S^4G data analysis pipeline and present measurements for a first set of galaxies, observed in both the cryogenic and warm mission phases of Spitzer. For every galaxy we tabulate the galaxy diameter, position angle, axial ratio, inclination at μ_(3.6μm)(AB) = 25:5, and 26:5 mag arcsec^(-2) (equivalent to ≈μ_B(AB) = 27:2 and 28:2 mag arcsec^(-2), respectively). These measurements will form the initial S^4G catalog of galaxy properties. We also measure the total magnitude and the azimuthally averaged radial profiles of ellipticity, position angle, surface brightness, and color. Finally, using the galaxy-fitting code GALFIT, we deconstruct each galaxy into its main constituent stellar components: the bulge/spheroid, disk, bar, and nuclear point source, where necessary. Together, these data products will provide a comprehensive and definitive catalog of stellar structures, mass, and properties of galaxies in the nearby universe and will enable a variety of scientific investigations, some of which are highlighted in this introductory S^4G survey paper.

Ketones and lactate "fuel" tumor growth and metastasis: Evidence that epithelial cancer cells use oxidative mitochondrial metabolism

Abstract: Previously, we proposed a new model for understanding the "Warburg effect" in tumor metabolism. In this scheme, cancer-associated fibroblasts undergo aerobic glycolysis and the resulting energy-rich metabolites are then transferred to epithelial cancer cells, where they enter the TCA cycle, resulting in high ATP production via oxidative phosphorylation. We have termed this new paradigm "The Reverse Warburg Effect." Here, we directly evaluate whether the end-products of aerobic glycolysis (3-hydroxy-butyrate and L-lactate) can stimulate tumor growth and metastasis, using MDA-MB-231 breast cancer xenografts as a model system. More specifically, we show that administration of 3-hydroxy-butyrate (a ketone body) increases tumor growth by ∼2.5-fold, without any measurable increases in tumor vascularization/angiogenesis. Both 3-hydroxy-butyrate and L-lactate functioned as chemo-attractants, stimulating the migration of epithelial cancer cells. Although L-lactate did not increase primary tumor growth, it stimulated the formation of lung metastases by ∼10-fold. Thus, we conclude that ketones and lactate fuel tumor growth and metastasis, providing functional evidence to support the "Reverse Warburg Effect". Moreover, we discuss the possibility that it may be unwise to use lactate-containing i.v. solutions (such as Lactated Ringer's or Hartmann's solution) in cancer patients, given the dramatic metastasis-promoting properties of L-lactate. Also, we provide evidence for the up-regulation of oxidative mitochondrial metabolism and the TCA cycle in human breast cancer cells in vivo, via an informatics analysis of the existing raw transcriptional profiles of epithelial breast cancer cells and adjacent stromal cells. Lastly, our findings may explain why diabetic patients have an increased incidence of cancer, due to increased ketone production, and a tendency towards autophagy/mitophagy in their adipose tissue.

Fault tolerance in distributed systems

Abstract: Fault tolerance is provided in a distributed system. The complexity of replicas and rollback requests are avoided; instead, a local failure in a component of a distributed system is tolerated. The local failure is tolerated by storing state related to a requested operation on the component, persisting that stored state in a data store, such as a relational database, asynchronously processing the operation request, and if a failure occurs, restarting the component using the stored state from the data store.

Towards a Rigorous Assessment of Systems Biology Models: The DREAM3 Challenges

Abstract: Background Systems biology has embraced computational modeling in response to the quantitative nature and increasing scale of contemporary data sets. The onslaught of data is accelerating as molecular profiling technology evolves. The Dialogue for Reverse Engineering Assessments and Methods (DREAM) is a community effort to catalyze discussion about the design, application, and assessment of systems biology models through annual reverse-engineering challenges.

A comparison of join algorithms for log processing in MaPreduce

Abstract: The MapReduce framework is increasingly being used to analyze large volumes of data. One important type of data analysis done with MapReduce is log processing, in which a click-stream or an event log is filtered, aggregated, or mined for patterns. As part of this analysis, the log often needs to be joined with reference data such as information about users. Although there have been many studies examining join algorithms in parallel and distributed DBMSs, the MapReduce framework is cumbersome for joins. MapReduce programmers often use simple but inefficient algorithms to perform joins. In this paper, we describe crucial implementation details of a number of well-known join strategies in MapReduce, and present a comprehensive experimental comparison of these join techniques on a 100-node Hadoop cluster. Our results provide insights that are unique to the MapReduce platform and offer guidance on when to use a particular join algorithm on this platform.

Coresets, sparse greedy approximation, and the Frank-Wolfe algorithm

Abstract: The problem of maximizing a concave function f(x) in the unit simplex Δ can be solved approximately by a simple greedy algorithm. For given k, the algorithm can find a point x(k) on a k-dimensional face of Δ, such that f(x(k) ≥ f(xa) − O(1/k). Here f(xa) is the maximum value of f in Δ, and the constant factor depends on f. This algorithm and analysis were known before, and related to problems of statistics and machine learning, such as boosting, regression, and density mixture estimation. In other work, coming from computational geometry, the existence of ϵ-coresets was shown for the minimum enclosing ball problem by means of a simple greedy algorithm. Similar greedy algorithms, which are special cases of the Frank-Wolfe algorithm, were described for other enclosure problems. Here these results are tied together, stronger convergence results are reviewed, and several coreset bounds are generalized or strengthened.