Um middleware para computação paralela em clusters de multicores

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Um método de otimização da relação desempenho/consumo de energia para arquiteturas multi-cores heterogêneas em FPGA

Devido às tendências de crescimento da quantidade de dados processados e a crescente necessidade por computação de alto desempenho, mudanças significativas estão acontecendo no projeto de arquiteturas de computadores. Com isso, tem-se migrado do paradigma sequencial para o paralelo, com centenas ou milhares de núcleos de processamento em um mesmo chip. Dentro desse contexto, o gerenciamento de energia torna-se cada vez mais importante, principalmente em sistemas embarcados, que geralmente são alimentados por baterias. De acordo com a Lei de Moore, o desempenho de um processador dobra a cada 18 meses, porém a capacidade das baterias dobra somente a cada 10 anos. Esta situação provoca uma enorme lacuna, que pode ser amenizada com a utilização de arquiteturas multi-cores heterogêneas. Um desafio fundamental que permanece em aberto para estas arquiteturas é realizar a integração entre desenvolvimento de código embarcado, escalonamento e hardware para gerenciamento de energia. O objetivo geral deste trabalho de doutorado é investigar técnicas para otimização da relação desempenho/consumo de energia em arquiteturas multi-cores heterogêneas single-ISA implementadas em FPGA. Nesse sentido, buscou-se por soluções que obtivessem o melhor desempenho possível a um consumo de energia ótimo. Isto foi feito por meio da combinação de mineração de dados para a análise de softwares baseados em threads aliadas às técnicas tradicionais para gerenciamento de energia, como way-shutdown dinâmico, e uma nova política de escalonamento heterogeneity-aware. Como principais contribuições pode-se citar a combinação de técnicas de gerenciamento de energia em diversos níveis como o nível do hardware, do escalonamento e da compilação; e uma política de escalonamento integrada com uma arquitetura multi-core heterogênea em relação ao tamanho da memória cache L1.

A framework for memory contention analysis in multi-core platforms

The last decade has witnessed a major shift towards the deployment of embedded applications on multi-core platforms. However, real-time applications have not been able to fully benefit from this transition, as the computational gains offered by multi-cores are often offset by performance degradation due to shared resources, such as main memory. To efficiently use multi-core platforms for real-time systems, it is hence essential to tightly bound the interference when accessing shared resources. Although there has been much recent work in this area, a remaining key problem is to address the diversity of memory arbiters in the analysis to make it applicable to a wide range of systems. This work handles diverse arbiters by proposing a general framework to compute the maximum interference caused by the shared memory bus and its impact on the execution time of the tasks running on the cores, considering different bus arbiters. Our novel approach clearly demarcates the arbiter-dependent and independent stages in the analysis of these upper bounds. The arbiter-dependent phase takes the arbiter and the task memory-traffic pattern as inputs and produces a model of the availability of the bus to a given task. Then, based on the availability of the bus, the arbiter-independent phase determines the worst-case request-release scenario that maximizes the interference experienced by the tasks due to the contention for the bus. We show that the framework addresses the diversity problem by applying it to a memory bus shared by a fixed-priority arbiter, a time-division multiplexing (TDM) arbiter, and an unspecified work-conserving arbiter using applications from the MediaBench test suite. We also experimentally evaluate the quality of the analysis by comparison with a state-of-the-art TDM analysis approach and consistently showing a considerable reduction in maximum interference.

Age determination of North Atlantic sediment cores

Available overwash records from coastal barrier systems document significant variability in North Atlantic hurricane activity during the late Holocene. The same climate forcings that may have controlled cyclone activity over this interval (e.g., the West African Monsoon, El Niño-Southern Oscillation (ENSO)) show abrupt changes around 6000 yrs B.P., but most coastal sedimentary records do not span this time period. Establishing longer records is essential for understanding mid-Holocene patterns of storminess and their climatic drivers, which will lead to better forecasting of how climate change over the next century may affect tropical cyclone frequency and intensity. Storms are thought to be an important mechanism for transporting coarse sediment from shallow carbonate platforms to the deep-sea, and bank-edge sediments may offer an unexplored archive of long-term hurricane activity. Here, we develop this new approach, reconstructing more than 7000 years of North Atlantic hurricane variability using coarse-grained deposits in sediment cores from the leeward margin of the Great Bahama Bank. High energy event layers within the resulting archive are (1) broadly correlated throughout an offbank transect of multi-cores, (2) closely matched with historic hurricane events, and (3) synchronous with previous intervals of heightened North Atlantic hurricane activity in overwash reconstructions from Puerto Rico and elsewhere in the Bahamas. Lower storm frequency prior to 4400 yrs B.P. in our records suggests that precession and increased NH summer insolation may have greatly limited hurricane potential intensity, outweighing weakened ENSO and a stronger West African Monsoon-factors thought to be favorable for hurricane development.

Preemption-light multiprocessor scheduling of sporadic tasks with high utilisation bound

Known algorithms capable of scheduling implicit-deadline sporadic tasks over identical processors at up to 100% utilisation invariably involve numerous preemptions and migrations. To the challenge of devising a scheduling scheme with as few preemptions and migrations as possible, for a given guaranteed utilisation bound, we respond with the algorithm NPS-F. It is configurable with a parameter, trading off guaranteed schedulable utilisation (up to 100%) vs preemptions. For any possible configuration, NPS-F introduces fewer preemptions than any other known algorithm matching its utilisation bound. A clustered variant of the algorithm, for systems made of multicore chips, eliminates (costly) off-chip task migrations, by dividing processors into disjoint clusters, formed by cores on the same chip (with the cluster size being a parameter). Clusters are independently scheduled (each, using non-clustered NPS-F). The utilisation bound is only moderately affected. We also formulate an important extension (applicable to both clustered and non-clustered NPS-F) which optimises the supply of processing time to executing tasks and makes it more granular. This reduces processing capacity requirements for schedulability without increasing preemptions.

Suporte à computação paralela e distribuída em Java: distribuição e execução de trabalho

Nos últimos anos começaram a ser vulgares os computadores dotados de multiprocessadores e multi-cores. De modo a aproveitar eficientemente as novas características desse hardware começaram a surgir ferramentas para facilitar o desenvolvimento de software paralelo, através de linguagens e frameworks, adaptadas a diferentes linguagens. Com a grande difusão de redes de alta velocidade, tal como Gigabit Ethernet e a última geração de redes Wi-Fi, abre-se a oportunidade de, além de paralelizar o processamento entre processadores e cores, poder em simultâneo paralelizá-lo entre máquinas diferentes. Ao modelo que permite paralelizar processamento localmente e em simultâneo distribuí-lo para máquinas que também têm capacidade de o paralelizar, chamou-se “modelo paralelo distribuído”. Nesta dissertação foram analisadas técnicas e ferramentas utilizadas para fazer programação paralela e o trabalho que está feito dentro da área de programação paralela e distribuída. Tendo estes dois factores em consideração foi proposta uma framework que tenta aplicar a simplicidade da programação paralela ao conceito paralelo distribuído. A proposta baseia-se na disponibilização de uma framework em Java com uma interface de programação simples, de fácil aprendizagem e legibilidade que, de forma transparente, é capaz de paralelizar e distribuir o processamento. Apesar de simples, existiu um esforço para a tornar configurável de forma a adaptar-se ao máximo de situações possível. Nesta dissertação serão exploradas especialmente as questões relativas à execução e distribuição de trabalho, e a forma como o código é enviado de forma automática pela rede, para outros nós cooperantes, evitando assim a instalação manual das aplicações em todos os nós da rede. Para confirmar a validade deste conceito e das ideias defendidas nesta dissertação foi implementada esta framework à qual se chamou DPF4j (Distributed Parallel Framework for JAVA) e foram feitos testes e retiradas métricas para verificar a existência de ganhos de performance em relação às soluções já existentes.

Sistema de ficheiros transaccional sobre FUSE

Dissertação apresentada na Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa para a obtenção do Grau de Mestre em Engenharia Informática

Parallel improved Schnorr-Euchner enumeration SE++ for the CVP and SVP

The Closest Vector Problem (CVP) and the Shortest Vector Problem (SVP) are prime problems in lattice-based cryptanalysis, since they underpin the security of many lattice-based cryptosystems. Despite the importance of these problems, there are only a few CVP-solvers publicly available, and their scalability was never studied. This paper presents a scalable implementation of an enumeration-based CVP-solver for multi-cores, which can be easily adapted to solve the SVP. In particular, it achieves super-linear speedups in some instances on up to 8 cores and almost linear speedups on 16 cores when solving the CVP on a 50-dimensional lattice. Our results show that enumeration-based CVP-solvers can be parallelized as effectively as enumeration-based solvers for the SVP, based on a comparison with a state of the art SVP-solver. In addition, we show that we can optimize the SVP variant of our solver in such a way that it becomes 35%-60% faster than the fastest enumeration-based SVP-solver to date.

A novel strategy for building interoperable MPI environment in heterogeneous high performance systems

Breakthrough advances in microprocessor technology and efficient power management have altered the course of development of processors with the emergence of multi-core processor technology, in order to bring higher level of processing. The utilization of many-core technology has boosted computing power provided by cluster of workstations or SMPs, providing large computational power at an affordable cost using solely commodity components. Different implementations of message-passing libraries and system softwares (including Operating Systems) are installed in such cluster and multi-cluster computing systems. In order to guarantee correct execution of message-passing parallel applications in a computing environment other than that originally the parallel application was developed, review of the application code is needed. In this paper, a hybrid communication interfacing strategy is proposed, to execute a parallel application in a group of computing nodes belonging to different clusters or multi-clusters (computing systems may be running different operating systems and MPI implementations), interconnected with public or private IP addresses, and responding interchangeably to user execution requests. Experimental results demonstrate the feasibility of this proposed strategy and its effectiveness, through the execution of benchmarking parallel applications.

A parallel method for impulsive image noise removal on hybrid CPU/GPU systems

A parallel algorithm for image noise removal is proposed. The algorithm is based on peer group concept and uses a fuzzy metric. An optimization study on the use of the CUDA platform to remove impulsive noise using this algorithm is presented. Moreover, an implementation of the algorithm on multi-core platforms using OpenMP is presented. Performance is evaluated in terms of execution time and a comparison of the implementation parallelised in multi-core, GPUs and the combination of both is conducted. A performance analysis with large images is conducted in order to identify the amount of pixels to allocate in the CPU and GPU. The observed time shows that both devices must have work to do, leaving the most to the GPU. Results show that parallel implementations of denoising filters on GPUs and multi-cores are very advisable, and they open the door to use such algorithms for real-time processing.

High performance and energy-efficient instruction cache design and optimisation for ultra-low-power multi-core clusters

High Energy efficiency and high performance are the key regiments for Internet of Things (IoT) end-nodes. Exploiting cluster of multiple programmable processors has recently emerged as a suitable solution to address this challenge. However, one of the main bottlenecks for multi-core architectures is the instruction cache. While private caches fall into data replication and wasting area, fully shared caches lack scalability and form a bottleneck for the operating frequency. Hence we propose a hybrid solution where a larger shared cache (L1.5) is shared by multiple cores connected through a low-latency interconnect to small private caches (L1). However, it is still limited by large capacity miss with a small L1. Thus, we propose a sequential prefetch from L1 to L1.5 to improve the performance with little area overhead. Moreover, to cut the critical path for better timing, we optimized the core instruction fetch stage with non-blocking transfer by adopting a 4 x 32-bit ring buffer FIFO and adding a pipeline for the conditional branch. We present a detailed comparison of different instruction cache architectures' performance and energy efficiency recently proposed for Parallel Ultra-Low-Power clusters. On average, when executing a set of real-life IoT applications, our two-level cache improves the performance by up to 20% and loses 7% energy efficiency with respect to the private cache. Compared to a shared cache system, it improves performance by up to 17% and keeps the same energy efficiency. In the end, up to 20% timing (maximum frequency) improvement and software control enable the two-level instruction cache with prefetch adapt to various battery-powered usage cases to balance high performance and energy efficiency.

The clusters Abell 222 and Abell 223: a multi-wavelength view

Context. The Abell 222 and 223 clusters are located at an average redshift z similar to 0.21 and are separated by 0.26 deg. Signatures of mergers have been previously found in these clusters, both in X-rays and at optical wavelengths, thus motivating our study. In X-rays, they are relatively bright, and Abell 223 shows a double structure. A filament has also been detected between the clusters both at optical and X-ray wavelengths. Aims. We analyse the optical properties of these two clusters based on deep imaging in two bands, derive their galaxy luminosity functions (GLFs) and correlate these properties with X-ray characteristics derived from XMM-Newton data. Methods. The optical part of our study is based on archive images obtained with the CFHT Megaprime/Megacam camera, covering a total region of about 1 deg(2), or 12.3 x 12.3 Mpc(2) at a redshift of 0.21. The X-ray analysis is based on archive XMM-Newton images. Results. The GLFs of Abell 222 in the g' and r' bands are well fit by a Schechter function; the GLF is steeper in r' than in g'. For Abell 223, the GLFs in both bands require a second component at bright magnitudes, added to a Schechter function; they are similar in both bands. The Serna & Gerbal method allows to separate well the two clusters. No obvious filamentary structures are detected at very large scales around the clusters, but a third cluster at the same redshift, Abell 209, is located at a projected distance of 19.2 Mpc. X-ray temperature and metallicity maps reveal that the temperature and metallicity of the X-ray gas are quite homogeneous in Abell 222, while they are very perturbed in Abell 223. Conclusions. The Abell 222/Abell 223 system is complex. The two clusters that form this structure present very different dynamical states. Abell 222 is a smaller, less massive and almost isothermal cluster. On the other hand, Abell 223 is more massive and has most probably been crossed by a subcluster on its way to the northeast. As a consequence, the temperature distribution is very inhomogeneous. Signs of recent interactions are also detected in the optical data where this cluster shows a ""perturbed"" GLF. In summary, the multiwavelength analyses of Abell 222 and Abell 223 are used to investigate the connection between the ICM and the cluster galaxy properties in an interacting system.

Landscape of intelligent cores: An autonomic multi-agent approach for space application

Space applications are challenged by the reliability of parallel computing systems (FPGAs) employed in space crafts due to Single-Event Upsets. The work reported in this paper aims to achieve self-managing systems which are reliable for space applications by applying autonomic computing constructs to parallel computing systems. A novel technique, 'Swarm-Array Computing' inspired by swarm robotics, and built on the foundations of autonomic and parallel computing is proposed as a path to achieve autonomy. The constitution of swarm-array computing comprising for constituents, namely the computing system, the problem / task, the swarm and the landscape is considered. Three approaches that bind these constituents together are proposed. The feasibility of one among the three proposed approaches is validated on the SeSAm multi-agent simulator and landscapes representing the computing space and problem are generated using the MATLAB.

Plant-symbiotic fungi as chemical engineers: multi-genome analysis of the Clavicipitaceae reveals dynamics of alkaloid Loci

The fungal family Clavicipitaceae includes plant symbionts and parasites that produce several psychoactive and bioprotective alkaloids. The family includes grass symbionts in the epichloae clade (Epichloë and Neotyphodium species), which are extraordinarily diverse both in their host interactions and in their alkaloid profiles. Epichloae produce alkaloids of four distinct classes, all of which deter insects, and some—including the infamous ergot alkaloids—have potent effects on mammals. The exceptional chemotypic diversity of the epichloae may relate to their broad range of host interactions, whereby some are pathogenic and contagious, others are mutualistic and vertically transmitted (seed-borne), and still others vary in pathogenic or mutualistic behavior. We profiled the alkaloids and sequenced the genomes of 10 epichloae, three ergot fungi (Claviceps species), a morning-glory symbiont (Periglandula ipomoeae), and a bamboo pathogen (Aciculosporium take), and compared the gene clusters for four classes of alkaloids. Results indicated a strong tendency for alkaloid loci to have conserved cores that specify the skeleton structures and peripheral genes that determine chemical variations that are known to affect their pharmacological specificities. Generally, gene locations in cluster peripheries positioned them near to transposon-derived, AT-rich repeat blocks, which were probably involved in gene losses, duplications, and neofunctionalizations. The alkaloid loci in the epichloae had unusual structures riddled with large, complex, and dynamic repeat blocks. This feature was not reflective of overall differences in repeat contents in the genomes, nor was it characteristic of most other specialized metabolism loci. The organization and dynamics of alkaloid loci and abundant repeat blocks in the epichloae suggested that these fungi are under selection for alkaloid diversification. We suggest that such selection is related to the variable life histories of the epichloae, their protective roles as symbionts, and their associations with the highly speciose and ecologically diverse cool-season grasses.

GALAXY EVOLUTION IN A COMPLEX ENVIRONMENT: A MULTI-WAVELENGTH STUDY OF HCG 7

The environment where galaxies are found heavily influences their evolution. Close groupings, like the ones in the cores of galaxy clusters or compact groups, evolve in ways far more dramatic than their isolated counterparts. We have conducted a multi-wavelength study of Hickson Compact Group 7 (HCG 7), consisting of four giant galaxies: three spirals and one lenticular. We use Hubble Space Telescope (HST) imaging to identify and characterize the young and old star cluster populations. We find young massive clusters (YMCs) mostly in the three spirals, while the lenticular features a large, unimodal population of globular clusters (GCs) but no detectable clusters with ages less than a few Gyr. The spatial and approximate age distributions of the similar to 300 YMCs and similar to 150 GCs thus hint at a regular star formation history in the group over a Hubble time. While at first glance the HST data show the galaxies as undisturbed, our deep ground-based, wide-field imaging that extends the HST coverage reveals faint signatures of stellar material in the intragroup medium (IGM). We do not, however, detect the IGM in H I or Chandra X-ray observations, signatures that would be expected to arise from major mergers. Despite this fact, we find that the H I gas content of the individual galaxies and the group as a whole are a third of the expected abundance. The appearance of quiescence is challenged by spectroscopy that reveals an intense ionization continuum in one galaxy nucleus, and post-burst characteristics in another. Our spectroscopic survey of dwarf galaxy members yields a single dwarf elliptical galaxy in an apparent stellar tidal feature. Based on all this information, we suggest an evolutionary scenario for HCG 7, whereby the galaxies convert most of their available gas into stars without the influence of major mergers and ultimately result in a dry merger. As the conditions governing compact groups are reminiscent of galaxies at intermediate redshift, we propose that HCGs are appropriate for studying galaxy evolution at z similar to 1-2.