Processing Element

Artificial neural networks (ANNs) are biologically inspired computer programs designed to simulate the way in which the human brain processes information. ANNs gather their knowledge by detecting the patterns and relationships in data and learn (or are trained) through experience, not from programming. An ANN is formed from hundreds of single units, artificial neurons or processing elements (PE), connected with coefficients (weights), which constitute the neural structure and are organised in layers. The power of neural computations comes from connecting neurons in a network. Each PE has weighted inputs, transfer function and one output. The behavior of a neural network is determined by the transfer functions of its neurons, by the learning rule, and by the architecture itself. The weights are the adjustable parameters and, in that sense, a neural network is a parameterized system. The weighed sum of the inputs constitutes the activation of the neuron. The activation signal is passed through transfer function to produce a single output of the neuron. Transfer function introduces non-linearity to the network. During training, the inter-unit connections are optimized until the error in predictions is minimized and the network reaches the specified level of accuracy. Once the network is trained and tested it can be given new input information to predict the output. Many types of neural networks have been designed already and new ones are invented every week but all can be described by the transfer functions of their neurons, by the learning rule, and by the connection formula. ANN represents a promising modeling technique, especially for data sets having non-linear relationships which are frequently encountered in pharmaceutical processes. In terms of model specification, artificial neural networks require no knowledge of the data source but, since they often contain many weights that must be estimated, they require large training sets. In addition, ANNs can combine and incorporate both literature-based and experimental data to solve problems. The various applications of ANNs can be summarised into classification or pattern recognition, prediction and modeling. Supervised associating networks can be applied in pharmaceutical fields as an alternative to conventional response surface methodology. Unsupervised feature-extracting networks represent an alternative to principal component analysis. Non-adaptive unsupervised networks are able to reconstruct their patterns when presented with noisy samples and can be used for image recognition. The potential applications of ANN methodology in the pharmaceutical sciences range from interpretation of analytical data, drug and dosage form design through biopharmacy to clinical pharmacy.

New abundances for neutron-capture (n-capture) elements in a large sample of metal-poor giants from the Bond survey are presented. The spectra were acquired with the KPNO 4-m echelle and coudé feed spectrographs, and have been analyzed using LTE fine-analysis techniques with both line analysis and spectral synthesis. Abundances of eight n-capture elements (Sr, Y, Zr, Ba, La, Nd, Eu, Dy) in 43 stars have been derived from blue (λλ4070-4710Å, R∼20,000, S/N ratio∼100-200) echelle spectra and red (λλ6100-6180Å, R∼22,000, S/N ratio∼100-200) coudé spectra, and the abundance of Ba only has been derived from the red spectra for an additional 27 stars.

We have used the Ultra-Violet Echelle Spectrograph (UVES) on Kueyen (UT2) of the VLT to take spectra of 15 individual red giant stars in the centers of four nearby dwarf spheroidal galaxies: Sculptor, Fornax, Carina and Leo I. We measure the abundance variations of numerous elements in these low mass stars with a range of ages (1−15 Gyr old). This means that we can effectively measure the chemical evolution of these galaxies with time.

Nervous systems are information processing networks that evolved by natural selection, whereas very large scale integrated (VLSI) computer circuits have evolved by commercially driven technology development. Here we follow historic intuition that all physical information processing systems will share key organizational properties, such as modularity, that generally confer adaptivity of function. It has long been observed that modular VLSI circuits demonstrate an isometric scaling relationship between the number of processing elements and the number of connections, known as Rent's rule, which is related to the dimensionality of the circuit's interconnect topology and its logical capacity. We show that human brain structural networks, and the nervous system of the nematode C. elegans, also obey Rent's rule, and exhibit some degree of hierarchical modularity. We further show that the estimated Rent exponent of human brain networks, derived from MRI data, can explain the allometric scaling relations between gray and white matter volumes across a wide range of mammalian species, again suggesting that these principles of nervous system design are highly conserved. For each of these fractal modular networks, the dimensionality of the interconnect topology was greater than the 2 or 3 Euclidean dimensions of the space in which it was embedded. This relatively high complexity entailed extra cost in physical wiring: although all networks were economically or cost-efficiently wired they did not strictly minimize wiring costs. Artificial and biological information processing systems both may evolve to optimize a trade-off between physical cost and topological complexity, resulting in the emergence of homologous principles of economical, fractal and modular design across many different kinds of nervous and computational networks.

We have obtained high-resolution, high S/N near-UV-blue spectra of 22 very metalpoor stars ([Fe/H] < −2.5) with Subaru/HDS, and measured the abundances of elements from C to Th. The metallicity range of the observed stars is −3.2 < [Fe/H] < −2.4. As found by previous studies, the star-to-star scatter in the measured abundances of neutron-capture elements in these stars is very large, much greater than could be assigned to observational errors, and in comparison with the relatively small scatter in the αand iron-peak elements. In spite of the large scatter in the ratios of the neutron-capture elements relative to iron, the abundance patterns of heavy neutroncapture elements (56 ≤ Z 72) are quite similar within our sample stars. The Ba/Eu ratios in the 11 very metal-poor stars in our sample in which both elements have been detected are nearly equal to that of the solar system r-process component. Moreover, the abundance patterns of the heavy neutron-capture elements (56 ≤ Z ≤ 70) in seven objects with clear enhancements of the neutron-capture elements are similar to that of the solar system r-process component. These results prove that heavy neutron-capture elements in these objects are primarily synthesized by the r-process.

Highly affect-laden memory intrusions are a feature of several psychological disorders with intrusive images of trauma especially associated with post-traumatic stress disorder (PTSD). The trauma film paradigm provides a prospective experimental tool for investigating analogue peri-traumatic cognitive mechanisms underlying intrusion development. We review several historical papers and some more recent key studies that have used the trauma film paradigm. A heuristic diagram is presented, designed to simplify predictions about analogue peri-traumatic processing and intrusion development, which can also be related to the processing elements of recent cognitive models of PTSD. Results show intrusions can be induced in the laboratory and their frequency amplified/attenuated in line with predictions. Successful manipulations include competing task type (visuospatial vs. verbal) and use of a cognitive coping strategy. Studies show that spontaneous peri-traumatic dissociation also affects intrusion frequency although attempts to manipulate dissociation have failed. It is hoped that further use of this paradigm may lead to prophylactic training for at risk groups and an improved understanding of intrusions across psychopathologies.

The study of human learning is complicated by the myriad of processing elements involved in conducting any behavioral task. In the case of visual perceptual learning, there has been significant controversy regarding the task processes that guide the formation of this learning. However, there is a developing consensus that top-down, task-related factors are required for such learning to take place. Here we challenge this idea by use of a novel procedure in which human participants, who were deprived of food and water, passively viewed visual stimuli while receiving occasional drops of water as rewards. Visual orientation stimuli, which were temporally paired with the liquid rewards, were viewed monocularly and rendered imperceptible by continuously flashing contour-rich patterns to the other eye. Results show that visual learning can be formed in human adults through stimulus-reward pairing in the absence of a task and without awareness of the stimulus presentation or reward contingencies.

In this paper we present the chemical abundance analysis from high resolution UVES spectra of seventeen bright giant stars of the Globular Cluster (GC) M 22. Results. We obtained an average iron abundance of [Fe/H] = -1.76 ± 0.02 (internal errors only) and an α enhancement of 0.36 ± 0.04 (internal errors only). Na and O, and Al and O follow the well known anti-correlation found in many other GCs. We identified two groups of stars with significantly different abundances of the s-process elements Y, Zr and Ba. The relative numbers of the two group members are very similar to the ratio of the stars in the two SGBs of M 22 recently found by Piotto (2009). Y and Ba abundances do not correlate with Na, O and Al. The s-element rich stars are also richer in iron and have higher Ca abundances. The results from high resolution spectra have been further confirmed by lower resolution GIRAFFE spectra of fourteen additional M 22 stars. GIRAFFE spectra show also that the Eu -a pure r-process element -abundance is not related to the iron content. We discuss the chemical abundance pattern of M 22 stars in the context of the multiple stellar populations in GC scenario.

The physical layer of most wireless protocols is traditionally implemented in custom hardware to satisfy the heavy computational requirements while keeping power consumption to a minimum. These implementations are time consuming to design and difficult to verify. A programmable hardware platform capable of supporting software implementations of the physical layer, or software defined radio, has a number of advantages. These include support for multiple protocols, faster time-to-market, higher chip volumes, and support for late implementation changes. The challenge is to achieve this without sacrificing power. In this paper, we present a design study for a fully programmable architecture, SODA, that supports software defined radio -a high-end signal processing application. Our design achieves high performance, energy efficiency, and programmability through a combination of features that include single-instruction multiple-data (SIMD) parallelism, and hardware optimized for 16bit computations. The basic processing element is an asymmetric processor consisting of a scalar and SIMD pipeline, and a set of distributed scratchpad memories that are fully managed in software. Results show that a four processor design is capable of meeting the throughput requirements of the W-CDMA and 802.11a protocols, while operating within the strict power constraints of a mobile terminal.

We summarize a selection process to identify red giants in the direction of the southern warp of the Galactic disk, employing VI_C photometry and multi-object spectroscopy. We also present results from follow-up high-resolution, high-S/N echelle spectroscopy of three field red giants, finding [Fe/H] values of about -0.5. The field stars, with Galactocentric distances estimated at 10 to 15 kpc, support the conclusion of Yong, Carney, & de Almeida (2005) that the Galactic metallicity gradient disappears beyond R_GC values of 10 to 12 kpc for the older stars and clusters of the outer disk. The field and cluster stars at such large distances show very similar abundance patterns, and, in particular, all show enhancements of the &quot;alpha&quot; elements O, Mg, Si, Ca, and Ti and the r-process element Eu. These results suggest that Type II supernovae have been significant contributors to star formation in the outer disk relative to Type Ia supernovae within the past few Gyrs. We also com...

HD 209458b is a Jovian planet orbiting around a star with characteristics similar to the Sun, which has repeated transits across the stellar disk. The orbital period is of 3.5 days, with a mean distance to the star of 0.045 AU. The peculiar characteristic of this system is that the planet shows a larger apparent size when observed in Lyα (1214-1217 Å) than at optical wavelengths. This discrepancy is thought to be due to the existence of an extended upper atmosphere and/or a cometary wake. In this work we present 3D numerical simulations of this system, and study the absorption produced by the extended atmosphere. This ``exosphere'' is modeled as a wind outflowing from the planet at the escape velocity (60 km s-1), which interacts with the stellar wind. Four runs where carried out with planetary mass loss rates in the 1.5 × 10-17 to 1.5 × 10-14 Msolar yr-1 range. By comparing the Lyα absorption predicted from the models with the observations, we determine an M˙p=(1.8+/-0.4)×10-16 Msolar yr-1 value for the mass loss rate from HD 209458b (a value which is of course dependent on the model assumptions, the error reflecting only the propagation of the observational uncertainties). This mass loss rate is consistent with the lower limit for M˙p (=1010 g s-1) determined by Vidal-Madjar and coworkers.

High-resolution spectroscopic observations of a hundred metal-poor Carbon and s-rich stars (CEMP-s) collected from the literature are compared with the theoretical nucleosynthesis models of asymptotic giant branch (AGB) presented in Paper I (M AGB ini = 1.3, 1.4, 1.5, 2 M ⊙ , −3.6 [Fe/H] −1.5). The s-process enhancement detected in these objects is associated to binary systems: the more massive companion evolved faster through the thermally pulsing AGB phase (TP-AGB), synthesising in the inner He-intershell the s-elements, which are partly dredged-up to the surface during the third dredge-up (TDU) episode. The secondary observed low mass companion became CEMP-s by mass transfer of C and s-rich material from the primary AGB. We analyse the light elements as C, N, O, Na and Mg, as well as the two s-process indicators, [hs/ls] (where ls = <Y, Zr> is the the light-s peak at N = 50 and hs = <La, Nd, Sm> the heavy-s peak at N = 82), and [Pb/hs]. We distinguish between CEMP-s with high s-process enhancement, [hs/Fe] 1.5 (CEMP-sII), and mild sprocess enhanced stars, [hs/Fe] < 1.5 (CEMP-sI). To interpret the observations, a range of s-process efficiencies at any given metallicity is necessary. This is confirmed by the high spread observed in [Pb/hs] (∼ 2 dex). A degeneration of solutions is found with some exceptions: most main-sequence CEMP-sII stars with low [Na/Fe] can only be interpreted with M AGB ini = 1.3 -1.4 M ⊙ . Giants having suffered the first dredge-up (FDU) need a dilution 1 dex (dil is defined as the mass of the convective envelope of the observed star, M obs ⋆ , over the material transferred from the AGB to the companion, M trans AGB ). Then, AGB models with higher AGB initial masses (M AGB ini = 1.5 -2 M ⊙ ) are adopted to interpret CEMP-sII giants. In general, solutions with AGB models in the mass range M AGB ini = 1.3 -2 M ⊙ and different dilution factors are found for CEMP-sI stars. About half of the CEMP-s stars with europium measurements show a high r-process enhancement (CEMP-s/r). The scenario for the origin of CEMP-s/r stars is a debated issue. We propose that the molecular cloud, from which the binary system formed, was previously enriched in r-process elements, most likely by local SN II pollution. This initial r-enrichment does not affect the s-process nucleosynthesis. However, for a high r-process enrichment ([r/Fe] ini = 2), the r-process contributions to solar La, Nd and Sm (30%, 40%, 70%) have to be considered. This increases the maximum [hs/ls] up to ∼ 0.3 dex. CEMP-s/r stars reflect this behaviour, showing higher [hs/ls] than observed in CEMP-s on average. Detailed analyses for individual stars will be provided in Paper III.

We introduce a 64-bit ANSI/IEEE Std 754-1985 floating point design of a hardware matrix multiplier optimized for FPGA implementations. A general block matrix multiplication algorithm, applicable for an arbitrary matrix size is proposed. The algorithm potentially enables optimum performance by exploiting the data locality and reusability incurred by the general matrix multiplication scheme and considering the limitations of the I/O bandwidth and the local storage volume. We implement a scalable linear array of processing elements (PE) supporting the proposed algorithm in the Xilinx Virtex II Pro technology. Synthesis results confirm a superior performance-area ratio compared to related recent works. Assuming the same FPGA chip, the same amount of local memory, and the same I/O bandwidth, our design outperforms related proposals by at least 1.7X and up to 18X consuming the least reconfigurable resources. A total of 39 PEs can be integrated into the xc2vp125-7 FPGA, reaching performance of, e.g., 15.6 GFLOPS with 1600 KB local memory and 400 MB/s external memory bandwidth.

P,*kt rbool~l b d4tM4• t, t c o Il•. 0 .ntotmatima estmnated t Image Understanding Architecture, Knowledge-Based Vision, AI Real-Time Computer Vision, Software Simulator, Parallel Processor IL PRICE CODE 17. SECURITY CLASSIFICATION 11. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. UMITATION OF ABSTRACT

We follow the chemical evolution of the Galaxy for elements from Ba to Eu, using an evolutionary model suitable to reproduce a large set of Galactic (local and non local) and extragalactic constraints. Input stellar yields for neutron-rich nuclei have been separated into their s-process and r-process components. The production of s-process elements in thermally pulsing asymptotic giant branch stars of low mass proceeds from the combined operation of two neutron sources:

The splicing processing defect in ATM intron 20 does not affect canonical splicing regulatory elements. a, Genomic localization and sequence analysis of the ATM cryptic exon. Gray boxes, normal ATM exons; thin line, normal ATM introns; black box, position of the indicated nucleotide sequence of the ATM intron. The deletion of 4 bp (GTAA) in intron 20 (boxed) causes the activation of the cryptic 65-bp exon (uppercase). The splice sites used by the cryptic exon are underlined (the 5′ splice site did not have the almost universal GU donor site, but its weak GC variant). Inclusion of the cryptic exon out of the reading frame is predicted to lead to the insertion of 23 amino-acid residues followed by the introduction of a stop codon. Because of the peculiar nucleotide arrangement, the same final sequence can be obtained if the deletion is shifted in the 3′ direction by one base (TAAG) or two bases (AAGT). We did not find the deletion in 50 normal chromosomes. b, RT-PCR products obtained from total RNA and spanning exons 20-21 showed an aberrant splicing event (ins 65) in a lymphoblastoid cell line derived from the affected individual (lane 1) that is absent in a lymphoblastoid cell line derived from a control (lane 2). M, size marker (1-kb ladder). Radioactive semiquantitative RT-PCR analysis showed that about 20-30% of the total mRNA obtained was the aberrantly spliced product (data not shown). We did not see aberrant exon inclusion in ATM mRNA from normal individuals in heart, liver, lung colon, spleen kidney and skin fibroblasts. c, Allele-specific analysis of the RT-PCR product amplified with primers designed to amplify sequence spanning exons 16 (ex 16 dir) and 22 (ex 22 rev) and separated on an agarose gel. As the other mutant alleles encode the 2250G→A splicing mutation 25 , which results in complete in-frame skipping of exon 16, the experiment indicates that the reduced abundance of misspliced product could not be attributed to leaky splicing. d, Lower abundance of the aberrant splicing allele results from nonsense-mediated decay. Puromycin treatment increases the relative amount of the product containing the 65-nt insertion from cell line of the affected individual. RT-PCR products were amplified with primers designed to form products spanning exons 19-22 (ex 19 dir-ex 22 rev) and separated on an agarose gel. Lane 1, no puromycin; lane 2, pre-incubation with puromycin. e, The 4-bp deletion causes aberrant inclusion of the cryptic exon. Left, hybrid minigenes containing the indicated ATM variants. pATM-∆ contains the patient's genomic sequences with the 4-nt deletion. The two adenines of the deletion in the pATM-CT minigene were replaced with the CT dinucleotide. Intronic ATM sequences (heavy lines) are inserted in a hybrid minigene system 11,26 . Black and shaded boxes, α-globin and fibronectin sequences, respectively. Transcription is driven by the minimal α-globin promoter and SV40 enhancer (small arrows at 5′ end). The primers used in the RT-PCR assay are represented by arrows. The splicing products corresponding to either complete removal of the intron or inclusion of the cryptic exon are indicated.

Developed for multimedia and game applications, as well as other numerically intensive workloads, the CELL processor provides support both for highly parallel codes, which have high computation and memory requirements, and for scalar codes, which require fast response time and a full-featured programming environment. This first generation CELL processor implements on a single chip a Power Architecture processor with two levels of cache, and eight attached streaming processors with their own local memories and globally coherent DMA engines. In addition to processor-level parallelism, each processing element has a Single Instruction Multiple Data (SIMD) unit that can process from 2 double precision floating points up to 16 bytes per instruction.

The recently-discovered polar codes are widely seen as a major breakthrough in coding theory. These codes achieve the capacity of many important channels under successive cancellation decoding. Motivated by the rapid progress in the theory of polar codes, we propose a family of architectures for efficient hardware implementation of successive cancellation decoders. We show that such decoders can be implemented with O(n) processing elements and O(n) memory elements, while providing constant throughput. We also propose a technique for overlapping the decoding of several consecutive codewords, thereby achieving a significant speed-up factor. We furthermore show that successive cancellation decoding can be implemented in the logarithmic domain, thereby eliminating the multiplication and division operations and greatly reducing the complexity of each processing element.

We present new spectroscopic observations for a sample of C(N)-type red giants. These objects belong to the class of Asymptotic Giant Branch stars, experiencing thermal instabilities in the He-burning shell (thermal pulses). Mixing episodes called third dredge-up enrich the photosphere with newly synthesized 12 C in the He-rich zone, and this is the source of the high observed ratio between carbon and oxygen (C/O ≥ 1 by number). Our spectroscopic abundance estimates confirm that, in agreement with the general understanding of the late evolutionary stages of low and intermediate mass stars, carbon enrichment is accompanied by the appearance of s-process elements in the photosphere. We discuss the details of the observations and of the derived abundances, focusing in particular on rubidium, a neutron-density sensitive element, and on the s-elements Sr, Y and Zr belonging to the first s-peak. The critical reaction branching at 85 Kr, which determines the relative enrichment of the studied species, is discussed.

Recent radiative lifetime measurements accurate to ±5% using laser-induced fluorescence (LIF) on 43 evenparity and 15 odd-parity levels of Ce ii have been combined with new branching fractions measured using a Fourier transform spectrometer (FTS) to determine transition probabilities for 921 lines of Ce ii. This improved laboratory data set has been used to determine a new solar photospheric Ce abundance, log ε = 1.61 ± 0.01 (σ = 0.06 from 45 lines), a value in excellent agreement with the recommended meteoritic abundance, log ε = 1.61 ± 0.02. Revised Ce abundances have also been derived for the r-process-rich metal-poor giant stars BD+17 • 3248, CS 22892-052, CS 31082-001, HD 115444, and HD 221170. Between 26 and 40 lines were used for determining the Ce abundance in these five stars, yielding a small statistical uncertainty of ±0.01 dex similar to the solar result. The relative abundances in the metal-poor stars of Ce and Eu, a nearly pure r-process element in the Sun, matches r-process-only model predictions for solar system material. This consistent match with small scatter over a wide range of stellar metallicities lends support to these predictions of elemental fractions. A companion paper includes an interpretation of these new precision abundance results for Ce as well as new abundance results and interpretation for Pr, Dy, and Tm.

The main component of the s process is produced by low mass stars (1.5 ≤ M/M ⊙ ≤ 3), when they climb for the second time the red giant branch and experience a series of He shell flashes called thermal pulses. During the relatively long period (10 5 yr) that elapses between two subsequent thermal pulses, a slow neutron flux is provided by the 13 C(α,n) 16 O reaction taking place within a thin 13 C pocket located in the He-rich and C-rich mantel of these stars. A second, marginal, neutron burst occurs during the thermal pulse and it is powered by the 22 Ne(α,n) 25 Mg reaction. We review the present status of the nucleosynthesis models of low mass AGB stars. The advance in the knowledge of the complex coupling between convective mixing and nuclear process, which allows the surface enrichment of C and s-process elements, is presented, together with the hypotheses concerning the physical mechanism driving the formation of the 13 C pocket. In order to illustrate the capabilities and the limits of the theory, an updated computation of a 2 M ⊙ stellar structure with solar chemical composition is reported. This model has been obtained by including a full nuclear network (from H up to Bi, at the termination point of the s-process path) into the stellar evolution code. The predicted modification of the surface composition occurring during the AGB evolution is shown. The new challenge of AGB modeling, namely the study of C-rich and s-rich very metal-poor stars, is discussed.

Media literacy education to promote health among youth involves them in a critical examination of media messages that promote risky behaviors and influence their perceptions and practices. Research on its effectiveness is in its infancy. Studies to date have been conducted with more or less rigor and achieved differing results, leaving many questions about effectiveness unanswered. To elucidate some of these questions, we conducted a systematic review of selected health-promoting media literacy education evaluation/research studies, guided by the following research question: What are the context and process elements of an effective health-promoting media literacy education intervention? Based on extensive analysis of 28 interventions, our findings provide a detailed picture of a small, 16-to 17-year (1990 to July 2006) body of important research, including citation information, health issue, target population/N/age, research design, intervention length and setting, concepts/skills taught, who delivered the intervention and ratings of effectiveness. The review provides a framework for organizing research about media literacy education which suggests that researchers should be more explicit about the media literacy core con-cepts/skills they are including in their interventions, and should more carefully address who delivered the intervention with what fidelity, in what setting, for how long and utilizing what pedagogical approach.

Some current and most future systems-on-chips use and will use network architectures/protocols to implement onchip communication. On-chip networks borrow features and design methods from those used in parallel computing clusters and computer system area networks. They differ from traditional networks because of larger on-chip wiring resources and flexibility, as well as constraints on area and energy consumption (in addition to performance requirements). In this paper, we analyze different routing schemes for packetized on-chip communication on a mesh network architecture, with particular emphasis on specific benefits and limitations of silicon VLSI implementations. A contention-look-ahead on-chip routing scheme is proposed. It reduces the network delay with significantly smaller buffer requirement. We further show that in the on-chip multiprocessor systems, both the instruction execution inside node processors, as well as data transaction between different processing elements, are greatly affected by the packetized dataflows that are transported on the on-chip networks. Different packetization schemes affect the performance and power consumption of multiprocessor systems. Our analysis is also quantified by the network/multiprocessor co-simulation benchmark results.

Hospitals and regional health authorities must set priorities in the face of resource constraints. Decision-makers seek practical ways to set priorities fairly in strategic planning, but find limited guidance from the literature. Very little has been reported from the perspective of Board members and senior managers about what criteria, processes and parameters of success they would use to set priorities fairly. We facilitated workshops for board members and senior leadership at three health care organizations to assist them in developing a strategy for fair priority setting. Workshop participants identified 8 priority setting criteria, 10 key priority setting process elements, and 6 parameters of success that they would use to set priorities in their organizations. Decision-makers in other organizations can draw lessons from these findings to enhance the fairness of their priority setting decision-making. Lessons learned in three workshops fill an important gap in the literature ab...

Many significant engineering and scientific problems involve optimization of some criteria over a combinatorial configuration space. The two methods most often used to solve these problems effectively-simulated annealing (SA) and genetic algorithms (GA)-do not easily lend themselves to massive parallel implementations. Simulated annealing is a naturally serial algorithm, while GA involves a selection process that requires global coordination. This paper introduces a new hybrid algorithm that inherits those aspects of GA that lend themselves to parallelization, and avoids serial bottle-necks of GA approaches by incorporating elements of SA to provide a completely parallel, easily scalable hybrid GA/SA method. This new method, called Genetic Simulated Annealing, does not require parallelization of any problem specific portions of a serial implementation-existing serial implementations can be incorporated as is. Results of a study on two difficult combinatorial optimization problems, a 100 city traveling salesperson problem and a 24 word, 12 bit error correcting code design problem, performed on a 16K PE MasPar MP-1, indicate advantages over previous parallel GA and SA approaches. One of the key results is that the performance of the algorithm scales up linearly with the increase of processing elements, a feature not demonstrated by any previous parallel GA or SA approaches, which enables the new algorithm to utilize massive parallel architecture with maximum effectiveness. Additionally, the algorithm does not require careful choice of control parameters, a significant advantage over SA and GA.

We study the slow neutron capture process (s process) in Asymptotic Giant Branch (AGB) stars using three different stellar evolutionary models computed for a 3 M ⊙ and solar metallicity star. First we investigate the formation and the efficiency of the main neutron source: the 13 C(α,n) 16 O reaction that occurs in radiative conditions. A tiny region rich in 13 C (the 13 C pocket) is created by proton captures on the abundant 12 C in the top layers of the He intershell, the zone between the H shell and the He shell. We parametrically vary the number of protons mixed from the envelope. For high local protons over 12 C number ratio, p/ 12 C ∼ > 0.3, most of the 13 C nuclei produced are further converted by proton capture to 14 N. Besides, 14 N nuclei represent a major neutron poison. We find

We present the abundance analysis of a sample of more than 120 red giants in the globular cluster (GC) NGC 1851, based on FLAMES spectra. We find a small but detectable metallicity spread. This spread is compatible with the presence of two different groups of stars with a metallicity difference of 0.06-0.08 dex, in agreement with earlier photometric studies. If stars are divided into these two groups according to their metallicity, both components show a Na-O anticorrelation (signature of a genuine GC nature) of moderate extension. The metal-poor stars are more concentrated than the metal-rich ones. We tentatively propose the hypothesis that NGC 1851 formed from a merger of two individual GCs with a slightly different Fe and α−element content, and possibly an age difference up to 1 Gyr. This is supported also by number ratios of stars on the split subgiant and on the bimodal horizontal branches. The distribution of n-capture process elements in the two components also supports the idea that the enrichment must have occurred in each of the structures separately, and not as a continuum of events in a single GC. The most probable explanation is that the proto-clusters formed into a (now dissolved) dwarf galaxy and later merged to produce the present GC.

Sounds provide us with useful information about our environment which complements that provided by other senses, but also poses specific processing problems. How does the auditory system distentangle sounds from different sound sources? And what is it that allows intermittent sound events from the same source to be associated with each other? Here we review findings from a wide range of studies using the auditory streaming paradigm in order to formulate a unified account of the processes underlying auditory perceptual organization. We present new computational modelling results which replicate responses in primary auditory cortex [Fishman, Y.I., Arezzo, J.C., Steinschneider, M., 2004. Auditory stream segregation in monkey auditory cortex: effects of frequency separation, presentation rate, and tone duration. J. Acoust. Soc. Am. 116, 1656-1670; Fishman, Y. I., Reser, D. H., Arezzo, J.C., Steinschneider, M., 2001. Neural correlates of auditory stream segregation in primary auditory cortex of the awake monkey. Hear. Res. 151, 167-187] to tone sequences. We also present the results of a perceptual experiment which confirm the bi-stable nature of auditory streaming, and the proposal that the gradual build-up of streaming may be an artefact of averaging across many subjects [Pressnitzer, D., Hupé, J. M., 2006. Temporal dynamics of auditory and visual bi-stability reveal common principles of perceptual organization. Curr. Biol. 16(13), 1351-1357.]. Finally we argue that in order to account for all of the experimental findings, computational models of auditory stream segregation require four basic processing elements; segregation, predictive modelling, competition and adaptation, and that it is the formation of effective predictive models which allows the system to keep track of different sound sources in a complex auditory environment.

I NFORMATION SYSTEMS STRATEGY PLANNING (ISSP) has been identified as vital to continuing organizational success and effective IS performance.'-3 However, although ISSP methodologies have been used for many years, organizations are still failing to deal effectively with untangling IS planning-related problems. This is mainly because many ISSP methodologies do not seem to realize that the IS-related problems result not merely from technological problems, but are also derived from paying too little attention to the interrelationship between IS and business and organizational factors." Research has shown that despite the growing number of theoretical frameworks for ISSP, practice still faces three major problems: ?? need for integration of ISSP and overall corporate ?? moderate practical utility of existing ISSP meth-?? limited management involvement and commitment to the ISSP activities4."

Extensible scientific visualization tools are often offered as data analysis tools. While images may be the goal of visualization, insight is the goal of analysis. Visualization tools often fail to reflect this fact both in functionality and in their user interfaces, which typically focus on graphics and programming concepts rather than on concepts more meaningful to end-user scientists. This paper presents a characterization which shows how data visualization5ts into the broader process of scientific data analysis. We conducted an empirical sludy, observing scientists from several disciplines while they analyzed their own data. Examination of the observations exposed process elements outside conventional image viewing. For example, analysts queried for quantitative information, made a variety of comparisons, applied math, managed data, and kept records. The characterization of scientific data analysis reveals activity beyond that traditionally supported by computer. It offers an understanding which has the potential to be applied to many future designs, and suggests specific recommendations for improving the support of this important aspect of scientijic computing.

Efficient run-time mapping of tasks onto Multiprocessor System-on-Chip (MPSoC) is very challenging especially when new tasks of other applications are also required to be supported at run-time. In this paper, we present a number of communication-aware run-time mapping heuristics for the efficient mapping of multiple applications onto an MPSoC platform in which more than one task can be supported by each processing element (PE). The proposed mapping heuristics examine the available resources prior to recommending the adjacent communicating tasks on to the same PE. In addition, the proposed heuristics give priority to the tasks of an application in close proximity so as to further minimize the communication overhead. Our investigations show that the proposed heuristics are capable of alleviating Networkon-Chip (NoC) congestion bottlenecks when compared to existing alternatives. We map tasks of applications onto an 8 Â 8 NoC-based MPSoC to show that our mapping heuristics consistently leads to reduction in the total execution time, energy consumption, average channel load and latency. In particular, we show that energy savings can be up to 44% and average channel load is improved by 10% for some cases.

Artificial Neural Networks are proposed to model and predict electricity consumption of Turkey. Multi layer perceptron with backpropagation training algorithm is used as the neural network topology. Tangent-sigmoid and pure-linear transfer functions are selected in the hidden and output layer processing elements, respectively. These input-output network models are a result of relationships that exist among electricity consumption and several other socioeconomic variables. Electricity consumption is modeled as a function of economic indicators such as population, gross national product, imports and exports. It is also modeled using export-import ratio and time input only. Performance comparison among different models is made based on absolute and percentage mean square error. Electricity consumption of Turkey is predicted until 2027 using data from 1975 to 2006 along with other economic indicators. The results show that electricity consumption can be modeled using Artificial Neural Networks, and the models can be used to predict future electricity consumption.

Abundances for 27 elements in the very metal poor dwarf star HE 2148-1247 are presented, including many of the neutron capture elements. We establish that HE 2148-1247 is a very highly s-process enhanced star with anomalously high Eu as well, Eu/H ∼half Solar, demonstrating the large addition of heavy nuclei at [Fe/H] = −2.3 dex. Ba and La are enhanced by a somewhat larger factor and reach the solar abundance, while Pb significantly exceeds it, thus demonstrating the addition of substantial s-process material. Ba/Eu is ten times the solar r-process ratio but much less than that of the s-process, indicating a substantial r-process addition as well. C and N are also very highly enhanced. We have found that HE 2148-1247 is a radial velocity variable; it is probably a small amplitude long period binary. The C, N and the s-process element enhancements thus presumably were produced through mass transfer from a former AGB binary companion. The large enhancement of heavy r-nuclides also requires an additional source as this is far above any inventory in the ISM at such low [Fe/H]. We consider that the s-process material was added by mass transfer of a more massive companion during its thermally pulsating AGB phase and ending up as a white dwarf. We further hypothesize that accretion onto the white dwarf from the envelope of the star caused accretion induced collapse of the white dwarf, forming a neutron star, which then produced heavy r-nuclides and again contaminated its companion. This mechanism in a binary system can thus enhance the envelope of the lower mass star in s and r-process material sequentially. Through analysis of the neutron capture element abundances taken from the literature for a large sample of very metal poor stars, we demonstrate, as exemplified by HE 2148-1247, that mass transfer in a suitable binary can be very efficient in enhancing the heavy elements in a star; it appears to be capable of enhancing 1 Based on observations obtained at the W.M.

The visual attention mechanism, which is the way humans perform object recognition , was applied to the implementation of a high performance object recognition chip . Even though the previous chip achieved 50% gain of computational cost [2], it could recognize only one object in a frame so that it is not suitable for advanced multi-object recognition applications such as video surveillance, intelligent robots, and autonomous vehicle navigation .

Although mesh-connected computers are used almost exclusively for low-level local image processing, they are also suitable for higher level image processing tasks. We illustrate this by presenting new optimal (in the 0-notational sense) algorithms for computing several geometric properties of figures. For example, given a black/white picture stored one pixel per processing element in an n X n mesh-connected computer, we give 0(n) time algorithms for determining the extreme points of the convex hull of each component, for deciding if the convex hull of each component contains pixels that are not members of the component, for deciding if two sets of processors are linearly separable, for deciding if each component is convex, for determining the distance to the nearest neighboring component of each component, for determining internal distances in each component, for counting and marking minimal internal paths in each component, for computing the external diameter of each component, for solving the largest empty circle problem, for determining internal diameters of components without holes, and for solving the all-points farthest point problem. Previous meshconnected computer algorithms for these problems were either nonexistent or had worst case times of 0 (n 2). Since any serial computer has a best case time of 0(n 2) when processing an n X n image, our algorithms show that the mesh-connected computer provides significantly better solutions to these problems.

In a parallel discrete-event simulation (PDES) scheme, tasks are distributed among processing elements (PEs) whose progress is controlled by a synchronization scheme. For lattice systems with short-range interactions, the progress of the conservative PDES scheme is governed by the Kardar-Parisi-Zhang equation from the theory of nonequilibrium surface growth. Although the simulated (virtual) times of the PEs progress at a nonzero rate, their standard deviation (spread) diverges with the number of PEs, hindering efficient data collection. We show that weak random interactions among the PEs can make this spread nondivergent. The PEs then progress at a nonzero, near-uniform rate without requiring global synchronizations.

We present new abundance determinations of neutron-capture elements Ge, Zr, Os, Ir, and Pt in a sample of 11 metal-poor (À3:1 ½Fe /H À1:6) Galactic halo giant stars, based on Hubble Space Telescope UV and Keck I optical high-resolution spectroscopy. The stellar sample is dominated by r-process-rich stars such as the well-studied CS 22892À052 and BD +17 3248 but also includes the r-process-poor, bright giant HD 122563. Our results demonstrate that abundances of the third r-process peak elements Os, Ir, and Pt in these metal-poor halo stars are very well correlated among themselves and with the abundances of the canonical r-process element Eu (determined in other studies), thus arguing for a common origin or site for r-process nucleosynthesis of heavier (Z > 56) elements. However, the large (and correlated) scatters of [Eu, Os, Ir, Pt /Fe] suggest that the heaviest neutron-capture r-process elements are not formed in all supernovae. In contrast, the Ge abundances of all program stars track their Fe abundances, very well. An explosive process on iron peak nuclei (e.g., the -rich freezeout in supernovae), rather than neutron capture, appears to have been the dominant synthesis mechanism for this element at low metallicities: Ge abundances seem completely uncorrelated with Eu. The correlation (with very small scatter) of Ge and Fe abundances suggests that Ge must have been produced rather commonly in stars, even at early times in the Galaxy, over a wide range of metallicity. The Zr abundances show much the same behavior as Ge with (perhaps) somewhat more scatter, suggesting some variations in abundance with respect to Fe. The Zr abundances also do not vary cleanly with Eu abundances, indicating a synthesis origin different than that of heavier neutron-capture elements. Detailed abundance distributions for CS 22892À052 and BD +17 3248, combining the new elemental determinations for Os-Pt and recently published Nd and Ho measurements, show excellent agreement with the solar system r-process curve from the elements Ba to Pb. The lighter n-capture elements, including Ge, in general fall below the same solar system r-process curve that matches the heavier elements.

Considerable research has been conducted on various issues associated with HIV testing. However, rather than conceptualizing HIV testing as a dynamic process which consists of interrelated elements, this body of work has focused on discrete aspects of the HIV testing process. As an example of such research, studies which have examined HIV testing in terms of various behavioural and psychological outcomes are critically reviewed. Their limitations are attributed to their failure to account for all the elements involved in the HIV testing process that-singly and in dynamic combination could have produced the measured outcomes. It is contended that if research on HIV testing is to be of use in the development and improvement of HIV testing services, it should be able to identify and describe in detail the factors that might lead to various outcomes of testing. This requires an in-depth examination of all aspects of the HIV testing process and their interrelationships from the perspectives of those undergoing testing and those providing testing services.The principal process elements in HIV testing are described: existing research on these topics is critically reviewed: and recommendations are made for future research. The process elements of l-! IV testing are identified as making a decision to be tested: accessing testing services; test counselling: and waiting for the test result. Of these, most consideration is accorded to the HIV test counselling process. It is contended that research is needed which examines both clients' and counsellors' expectations, experiences of and satisfaction with HIV test counselling. Specific issues that could usefully be addressed by future research include the process of obtaining clients' informed consent for testing; the ways in which test results are conveyed to clients; the strategies used in HIV counselling to help clients avoid or reduce risk behaviours in the future: partner notification: and which professional groups are best placed to conduct HIV test counselling. Flintily, the question of which research methods might be suitable ['or process-based studies is considered. It is concluded that qualitative methods could be particularly appropriate as they are well-placed to chart in detail the varied aspects of the HIV testing process and their interrelationships.

Managing a Network-on-Chip (NoC) in an efficient way is a challenging task. To succeed, the operating system (OS) needs to be tuned to the capabilities and the needs of the NoC. Only by creating a tight interaction can we combine the necessary flexibility with the required efficiency. This paper illustrates such an interaction by detailing the management of communication resources in a system containing a packet-switched NoC and a closely integrated OS. Our NoC system is emulated by linking an FPGA to a PDA. We show that, with the right NoC support, the OS is able to optimize communication resource usage. Additionally, the OS is able to diminish or remove the interference between independent applications sharing a common NoC communication resource.

A large sample of carbon enhanced metal-poor stars enriched in s-process elements (CEMP-s) have been observed in the Galactic halo. These stars of low mass (M ∼ 0.9 M ⊙ ) are located on the main-sequence or the red giant phase, and do not undergo third dredge-up (TDU) episodes. The s-process enhancement is most plausibly due to accretion in a binary system from a more massive companion when on the asymptotic giant branch (AGB) phase (now a white dwarf). In order to interpret the spectroscopic observations, updated AGB models are needed to follow in detail the sprocess nucleosynthesis. We present nucleosynthesis calculations based on AGB stellar models obtained with FRANEC (Frascati Raphson-Newton Evolutionary Code) for low initial stellar masses and low metallicities. For a given metallicity, a wide spread in the abundances of the s-process elements is obtained by varying the amount of 13 C and its profile in the pocket, where the 13 C(α, n) 16 O reaction is the major neutron source, releasing neutrons in radiative conditions during the interpulse phase. We account also for the second neutron source 22 Ne(α, n) 25 Mg, partially activated during convective thermal pulses. We discuss the surface abundance of elements from carbon to bismuth, for AGB models of initial masses M = 1.3 -2 M ⊙ , low metallicities ([Fe/H] from −1 down to −3.6) and for different 13 C-pockets efficiencies. In particular we analyse the relative behaviour of the three s-process peaks: light-s (ls at magic neutron number N = 50), heavy-s (hs at N = 82) and lead (N = 126). Two s-process indicators, [hs/ls] and [Pb/hs], are needed in order to characterise the s-process distribution. In the online material, we provide a set of data tables with surface predictions. Our final goal is to provide a full set of theoretical models of low mass low metallicity s-process enhanced stars. In a forthcoming paper, we will test our results through a comparison with observations of CEMP-s stars.