The effective use of GPUs for accelerating applications depends on a number of factors including effective asynchronous use of heterogeneous resources, reducing data transfer between CPU and GPU, increasing occupancy of GPU kernels, overlapping data transfers with computations, reducing GPU idling and kernel optimizations. Overcoming these challenges require considerable effort on the part of the application developers. Most optimization strategies are often proposed and tuned specifically for individual applications. Message-driven executions with over-decomposition of tasks constitute an important model for parallel programming and provide multiple benefits including communication-computation overlap and reduced idling on resources. Charm++ is one such message-driven language which employs over decomposition of tasks, computation-communication overlap and a measurement-based load balancer to achieve high CPU utilization. This research has developed an adaptive runtime framework for efficient executions of Charm++ message-driven parallel applications on GPU systems. In the first part of our research, we have developed a runtime framework, G-Charm with the focus primarily on optimizing regular applications. At runtime, G-Charm automatically combines multiple small GPU tasks into a single larger kernel which reduces the number of kernel invocations while improving CUDA occupancy. G-Charm also enables reuse of existing data in GPU global memory, performs GPU memory management and dynamic scheduling of tasks across CPU and GPU in order to reduce idle time. In order to combine the partial results obtained from the computations performed on CPU and GPU, G-Charm allows the user to specify an operator using which the partial results are combined at runtime. We also perform compile time code generation to reduce programming overhead. For Cholesky factorization, a regular parallel application, G-Charm provides 14% improvement over a highly tuned implementation. In the second part of our research, we extended our runtime to overcome the challenges presented by irregular applications such as a periodic generation of tasks, irregular memory access patterns and varying workloads during application execution. We developed models for deciding the number of tasks that can be combined into a kernel based on the rate of task generation, and the GPU occupancy of the tasks. For irregular applications, data reuse results in uncoalesced GPU memory access. We evaluated the effect of altering the global memory access pattern in improving coalesced access. We’ve also developed adaptive methods for hybrid execution on CPU and GPU wherein we consider the varying workloads while scheduling tasks across the CPU and GPU. We demonstrate that our dynamic strategies result in 8-38% reduction in execution times for an N-body simulation application and a molecular dynamics application over the corresponding static strategies that are amenable for regular applications.

[pt] O presente trabalho de tese é destinado ao estudo do decaimento do méson D+ KS(pi)−(pi)+(pi)+ com dados coletados pelo experimento FOCUS do Fermilab. O estado final, composto por 4 mésons pseudo- escalares, pode ser alcançado através de uma considerável variedade de sub- estruturas ressonantes. O proposito da análise é encontrar quais são estes possíveis canais intermediários através de suas contribuições e fases relativas. Para tanto, é utilizado o formalismo de Análise de Amplitudes usando o chamado Modelo Isobárico. A dinâmica do decaimento é definida através de uma função que contém as características (formas funcionais das ressonâncias, distribuição angular, etc) dos canais que intervêm no decaimento e cujo domínio é um espaço o de fase determinado por 5 invariantes (devido às restrições de decaimento em quatro corpos sem spin). A função é então ajustada ao conjunto de dados coletados. Nos resultados, verificamos uma grande contribuição do méson vetor-axial a1(1260) (52%) seguido do vetor-axial K1(1400) (34%). Além disso, o modelo apresenta contribuição da ressonância (sigma) (cerca de 8%, vinda de a1(sigmapi)KS e sigmaKSpi) e uma razoável contribuição da ressonância escalar K− (14%). O estado K já foi visto em seu modo neutro em outros decaimentos de charme, porém ainda não em seu modo carregado. Não encontramos contribui ção significativa do decaimento direto em 4 corpos (não- ressonante). Esta tese vem se somar ao esforço no entendimento da dinâmica das intera ções fortes a baixas energias, que nos últimos anos tem ganhado da física de mésons charmosos uma importante contribuição. / [en] This thesis is devoted to the study of the decay D+ KS(pi)− (pi)+(pi)+ with data collected from the FOCUS experiment, at Fermilab. The final state composed of 4 pseudo-scalars can be produced through a number of resonant sub-structures. The purpose of this analysis is to find the contributing intermediate states by measuring their relative strenghts and phases. For that, the Amplitude Analysis formalism is used, with the so-called Isobar Model. The decay dynamics is described through a function which has the features of the contributing channels (functional forms of the resonances, angular distribution, etc) and which domain is a phase space determined by 5 invariants (due to the kinematical constraints of a 4- body spinless decay). The data sample is thus fitted to this function. Our results show a dominant contribution of the axial- vector meson a1(1260) (52%), followed by the K1(1400) axial-vector (34%). Moreover, the model presents a contribution from the sigma meson (about 8% as a1(sigmapi)KS and sigmaKSpi) and a significant contribution from the scalar k−. The k state has been reported in its neutral mode in other charm decays but not is its charged mode. We find no significant contribution from the direct 4-body decay (non-resonant). This work adds to the effort in the understanding of the strong-interaction dynamics at low energies, which in recent years have been receiving an important contribution from charm meson physics.

The emulsion target of the CHORUS detector was exposed to the wide-band neutrino beam of the CERN SPS between the years 1994 and 1997. In total about 130.000 neutrino interactions were located in the nuclear emulsion target and fully reconstructed. Detailed DATA and Monte Carlo (MC) comparisons were done in order to test reliability of the detector simulation. There is reasonable agreement between DATA and MC. The ratio of deep inelastic neutral-current (NC) to the deep inelastic (DIS) charged-current(CC) $nu_{mu}$ interactions was measured to be frac{NC_{dis}}{CC_{dis}}=0.350pm0.003$. This measurement was compared with the previous measurements. Based on three double charm decays found in NC interactions we measured the ratio of double charm cross-section in NC $nu_{mu}$ interactions to be [ frac{sigma (cbar{c}nu_mu)}{sigma_{NC}}=(3.37^{+3.06}_{-2.51}(stat.)pm 0.51(syst.))times 10^{-3}. ] One double charm decay has been observed in CC $nu_{mu}$ interactions the upper limit on associated charm production in $nu_{mu}$ CC interaction has been found to be [ frac{sigma cbar{c}mu^{-})}{sigma_{CC}}&lt / 9.69 times 10^{-4}. ] at 90% C.L.

Submitted by Edileide Reis (leyde-landy@hotmail.com) on 2013-05-06T14:25:40Z No. of bitstreams: 2 Araujo, Miguel de Parte 2.pdf: 2378843 bytes, checksum: 9c14cb3f47970fb1ac9aba612a8b258d (MD5) Araujo, Miguel de Parte 1.pdf: 889851 bytes, checksum: 8ff83550a38f6edfebb1571f896fc792 (MD5) / Approved for entry into archive by Maria Auxiliadora Lopes(silopes@ufba.br) on 2013-05-17T13:29:39Z (GMT) No. of bitstreams: 2 Araujo, Miguel de Parte 2.pdf: 2378843 bytes, checksum: 9c14cb3f47970fb1ac9aba612a8b258d (MD5) Araujo, Miguel de Parte 1.pdf: 889851 bytes, checksum: 8ff83550a38f6edfebb1571f896fc792 (MD5) / Made available in DSpace on 2013-05-17T13:29:39Z (GMT). No. of bitstreams: 2 Araujo, Miguel de Parte 2.pdf: 2378843 bytes, checksum: 9c14cb3f47970fb1ac9aba612a8b258d (MD5) Araujo, Miguel de Parte 1.pdf: 889851 bytes, checksum: 8ff83550a38f6edfebb1571f896fc792 (MD5) Previous issue date: 2006 / As meditações que descortino na tese emergem das in-tensidades de minhas inquietudes e espantamentos diante dos fenômenos do existir e da ação de educar; apresentam ?Uma compreensão ontológico-policrômica da Sensibilidade e sua fruição no fenômeno do educar? numa mirada Filantropoética. Mirada que busca entrelaçar o elã do filosófico-antropológico e a cromaticidade do poético vislumbrando a polifonia e a policromia dos Sentidos pregnantes e anímicos da temática. Dessa forma, apresento a Sensibilidade (Sensus) como estado de dis-posição, de abertura vasta de nossos sensos perceptivos (afeccionais e noéticos), em que corpo e espírito coexistem, de modo co-implicado, para uma compreensão e uma vivenciação policrômicas dos Sentidos do existir e do co-existir humanos; como estado de despojamento do espírito inventivo e altaneiro, do pathos criante e co-movente, para a vivência do sentimento do mundo, do ser-sendo-com-os-outros, das in-tensidades da complexidade e das ambigüidades da condição humana. Nesse horizonte compreensivo, a Sensibilidade se constitui desde os fulcros magmáticos da Corporeidade, da Intuição, da Afetividade, do Mitopoético e da Razão-Sentido. Fulcros estruturantes que se enredam de modo recursivo e entrelaçado. Em seguida, apresento meditações que compreendem o fenômeno do educar como ação teóricovivencial que implica na fruição da Sensibilidade. Para tanto, me inspiro, tanto nas ressonâncias das vozes de diversos estudantes que emergem de escutas sobre o vivido/vivente, como no estofo das incursões teóricas. Assim, apresento meditações acerca do educar compreendendo-o como um rito vivo de iniciação que se traduz em processos in-tensivos de con-dução aos saberes e sentires que constituem o dinamismo do ethos; no cuidado com a inteireza androgínica do ser-sendo-com no advento dos valores humanos primordiais, dos Sentidos anímicos, do cuidado com a Sensibilidade. Por fim, como ?Arremates inconcludentes?, propugno que, ao primar pela fruição da Sensibilidade, a ação de educar se desdobra numa Pedagogia do encantamento como se-ducere que implica no jorrar do pathos criante, do admirante, do elã vital, do espírito altaneiro e audaz; que envida apaixonamento e entusiasmo, laços de simpatia e de empatia; que conduz a fruição dos Sentidos pregnantes e anímicos no desbordar dos feixes do arco-íris que alumbram e revelam as policromias da poeticidade do existir, do co-existir. / Salvador

Agrégation de l'enseignement supérieur, Orientation sciences / Thèse d'agrégation / info:eu-repo/semantics/nonPublished

[pt] A melhor teoria atual para descrever partículas elementares e suas interações, o Modelo Padrão, tem como peça fundamental a violação de carga-paridade (CP), a qual implica diferença no comportamento de partículas e anti-partículas. Por conta da violação de CP, o decaimento de um méson em um certo estado final pode então ser distinto do decaimento do anti- méson nesse estado final conjugado. Esse efeito foi extensivamente estudado e medido ao longo dos anos em decaimentos de mésons K e B, que envolvem os quarks strange e beauty respectivamente, mas apenas recentemente observado em decaimentos de mésons D, envolvendo o quark charm. As assimetrias de CP previstas no setor charmoso são bastante pequenas, então para que novas observações sejam feitas, grandes conjuntos de dados são necessários. Esta dissertação reporta a busca por violação de CP nos decaimentos Ds+ -> K-K+K+, suprimido por Cabibbo, e D+ -> K-K+K+, duplamente suprimido por Cabibbo. A busca por assimetrias locais nos espaços de fase dos decaimentos é realizada com uma técnica modelo-independente, usando dados obtidos em colisões pp de energia de centro-de-massa de 13 TeV, coletados pelo detector LHCb durante 2016—2018 (run 2 ). Também incluído está um estudo da sensibilidade do método a efeitos de violação de CP dadas as estatísticas atuais usando simulações. Nesta análise, dois decaimentos favorecidos por Cabibbo, Ds+ -> K-K+pi+ e D+ -> K-pi+pi+, são utilizados para nos assegurar que não somos sensíveis a assimetrias espúrias, de efeitos de produção ou detecção. / [en] The current best theory for describing elementary particles and their interactions, the Standard Model, has charge-parity (CP) violation as a fundamental feature, which implies that particles and antiparticles behave differently. Due to CP violation, the decay of a meson into a certain final state might then be distinct from the decay of the anti-meson to that charge conjugated final state. This effect has been extensively studied and measured over the years in the decays of K and B mesons, which involve the strange and beauty quarks respectively, but only recently observed in D meson decays, involving the charm quark. The predicted CP asymmetries in the charm sector are very small, so, in order to make new observations, large data samples are required. This dissertation reports the search for CP violation in the singly- Cabibbo-suppressed Ds+ -> K-K+K+ and doubly-Cabibbo-suppressed D+ -> K-K+K+ decays. The search for local asymmetries in the phase space of the decays is performed with a model-independent technique, using the data samples obtained in pp collisions with a centre-of-mass energy of 13 TeV, collected by the LHCb detector during 2016—2018 (run 2). Also included is a study of the method s sensitivity to CP violation effects in the current statistics using simulations. In this analysis, two Cabibbo-favoured decays, Ds+ -> K-K+pi+ and D+ -> K-pi+pi+, are used to ensure we are not sensitive to nuisance asymmetries, from production or detection effects.

[pt] O Modelo Padrão (SM) da física de partículas é atualmente a melhor teoria para descrever as interações entre partículas elementares e suas propriedades. A observação da violação do CP em diversos processos fracos no setor de quarks - permitindo uma distinção absoluta entre partículas e antipartículas - é bem explicada pelo SM (através do chamado Cabibbo-Kobayashi-Maskawa ansatz). A violação do CP é uma das condições necessárias para a bariogênese e pode ser a chave para explicar a assimetria matéria-antimatéria no universo. Esta dissertação apresenta a busca por violação de CP no decaimento D+ → π−π+K+ duplamente suprimido por Cabibbo (DCS), usando dados coletados pelo LHCb de 2016-2018 de colisões pp com uma energia de centro de massa de 13 TeV, correspondendo a uma luminosidade integrada de 5,6 fb1. O objetivo desta análise é a implementação de uma técnica independente de modelo para realizar uma comparação estatística entre o espaço de fase, denominado Dalitz Plot (DP), de partícula e antipartícula no canal de decaimento, buscando diferenças locais na distribuição de eventos entre os dois DPs. Primeiro, foi feito um processo de seleção para remover contribuições de background de outros decaimentos de charme, bem como uma análise multivariada utilizando algoritmos de aprendizado de máquina para reduzir os níveis de background combinatoriais. Após a seleção, foi obtida uma amostra final 6M de candidatos a sinal, que é hoje a maior amostra já obtida para um canal de decaimento D+ DCS, permitindo uma excelente sensibilidade para busca de VCP. Essa é a primeira busca de VCP no canal de decaimento estudado. Esta análise é realizada de forma cega, o que significa que não há resultado para a região do sinal nesta primeira etapa e, para garantir que não haja assimetrias espúrias, como efeitos de produção e detecção, foram realizados testes para a região de background, para o canal de controle D+ → K−π+π+ e para a amostra simulada de Monte Carlo. / [en] The Standard Model (SM) of particle physics is currently the best theory to describe the interactions between elementary particles and their properties. The observation of CP violation in a variety of weak processes in the quark sector — allowing an absolute distinction of particles and antiparticles — is well explained by the SM (through so-called Cabibbo-Kobayashi-Maskawa ansatz). CP violation is one of the necessary conditions for baryogenesis and might be the key to explain the matter-antimatter asymmetry in the universe. This dissertation presents the search for CP violation in the doubly Cabibbo suppressed (DCS) D+ → π−π+K+ decay, using data collected by LHCb from 2016-2018 of pp collisions with a centre of mass energy of 13 TeV, corresponding to an integrated luminosity of 5.6 fb1. The goal of this analysis is the implementation of a model-independent technique to perform a statistical comparison between the phase space, called Dalitz Plot (DP), of particle and antiparticle decay channel, searching for local differences in the distribution of events between the two DPs. First, a selection process was executed to remove background contributions from other charm decays, as well as a multivariate analysis using machine learning algorithms to reduce combinatorial background levels. After the selection, a final sample of ¨6M signal candidates was obtained, which is nowadays the largest sample ever obtained for a DCS D+ decay channel, allowing an outstanding sensitivity for CPV search. This is the first CPV search in the studied decay channel. This analysis is performed blinded, meaning that there is no actual result for the signal region at this first stage and in order to guarantee that there are no nuisance asymmetries, from production and detection effects, tests were performed using the background region, the control channel D+ → K−π+π+ and the Monte Carlo simulated sample.

NIL

<p>The Standard Model describes all elementary particles known today, but at larger energies it will have to be complemented with new particles and interactions. To be able to distinguish new physics at proton colliders such as LHC at CERN, it is essential to have an appropriate description of the colliding protons and their interactions. The study of the proton is important also in itself, to get a better understanding of the non-perturbative aspects of the strong interaction.</p><p>In paper I-IV of this thesis, a model for the non-perturbative dynamics of quarks and gluons is developed, based on quantum fluctuations in hadrons. The parton distributions of the proton are given by momentum fluctuations, with sea quark distributions generated by fluctuations into baryon-meson pairs. This model can reproduce proton structure function data, as well as measured asymmetries between up and down valence quark distributions and between the anti-up and anti-down sea. It provides an intrinsic charm quark component as indicated by data. It also predicts an asymmetry in the strange sea of the proton, which can explain the NuTeV anomaly first attributed to new physics beyond the Standard Model.</p><p>Charged Higgs bosons are predicted by several theories for new physics, including Supersymmetry. At proton colliders, the predicted dominant production mechanism is in association with top and bottom quarks. In paper V-VII, different contributions to this production are studied, and an algorithm is developed for combining the two dominant processes gb -> tH^+/- and gg -> tbH^+/-. The algorithm gives a smooth transition from small to large transverse momenta of the b-quark, which is important when the b-quark is observed. It also gives arguments for the choice of factorisation scale in the process.</p>

The Standard Model describes all elementary particles known today, but at larger energies it will have to be complemented with new particles and interactions. To be able to distinguish new physics at proton colliders such as LHC at CERN, it is essential to have an appropriate description of the colliding protons and their interactions. The study of the proton is important also in itself, to get a better understanding of the non-perturbative aspects of the strong interaction. In paper I-IV of this thesis, a model for the non-perturbative dynamics of quarks and gluons is developed, based on quantum fluctuations in hadrons. The parton distributions of the proton are given by momentum fluctuations, with sea quark distributions generated by fluctuations into baryon-meson pairs. This model can reproduce proton structure function data, as well as measured asymmetries between up and down valence quark distributions and between the anti-up and anti-down sea. It provides an intrinsic charm quark component as indicated by data. It also predicts an asymmetry in the strange sea of the proton, which can explain the NuTeV anomaly first attributed to new physics beyond the Standard Model. Charged Higgs bosons are predicted by several theories for new physics, including Supersymmetry. At proton colliders, the predicted dominant production mechanism is in association with top and bottom quarks. In paper V-VII, different contributions to this production are studied, and an algorithm is developed for combining the two dominant processes gb -&gt; tH+/- and gg -&gt; tbH+/-. The algorithm gives a smooth transition from small to large transverse momenta of the b-quark, which is important when the b-quark is observed. It also gives arguments for the choice of factorisation scale in the process.