Solving a linear system of equations is a common task performed for most of the scientific experiments. Everyday, these applications increase their demands, pushing the computational capacity to the limits. At a certa...
详细信息
ISBN:
(纸本)9780769549347;9781467350266
Solving a linear system of equations is a common task performed for most of the scientific experiments. Everyday, these applications increase their demands, pushing the computational capacity to the limits. At a certain point, our systems, running common algorithms, become unable to process such massive amounts of information. the paper analyses and modifies the standard Gaussian reduction so that a huge array may be handled by almost any distributed system. Moreover, the paper focuses on providing a decent level of fault tolerance in order to comply with a real, heterogeneous distributed system and on becoming a solution for sensitive applications.
Simulation of quantum computers using classical computers is a computationally hard problem, requiring a huge amount of operations and storage. Grid systems are a good choice for simulating quantum algorithms, since t...
详细信息
ISBN:
(纸本)9780769539645
Simulation of quantum computers using classical computers is a computationally hard problem, requiring a huge amount of operations and storage. Grid systems are a good choice for simulating quantum algorithms, since they provide access to high performance computer clusters. In this paper we present the design of a message passing parallel version of the quantum computer simulator, QCL, deployed as a grid service. After describing the architecture of our grid service and the parallelization strategy for the general single qubit operator, we present the performance measurements for some test cases, showing the speedups obtained.
there is a wide range of applications of non-equidistant discretization of real signals. For instance, in computer graphics, Fourier analysis, identification and control theories, etc. they have the common ability to ...
详细信息
ISBN:
(纸本)9781479930357
there is a wide range of applications of non-equidistant discretization of real signals. For instance, in computer graphics, Fourier analysis, identification and control theories, etc. they have the common ability to describe dynamical systems as well. In this paper we provide a fast algorithm based on an existing mathematical model to compute a non-uniform grid for representing different types of signals. In order to do that we need new concepts for constructing an effective numerical solution. Additionally, two experiments are performed to investigate the accuracy of the method. Finally, we also present a parallel implementation in CUDA which can further improve the execution time.
Monitoring cloud resources is an essential part of cloud computing, and although some effort has been made in this direction, monitoring security parameters is still an open issue because of the large number and types...
详细信息
ISBN:
(纸本)9781509004614
Monitoring cloud resources is an essential part of cloud computing, and although some effort has been made in this direction, monitoring security parameters is still an open issue because of the large number and types of resources available in a multi tenant environment. In this paper we propose a distributed monitoring system that is able to monitor, filter and store security parameters of network hosts. the proposed system is decentralized, fault tolerant and can run scanning jobs in parallel, those three characteristics making it a perfect candidate for monitoring resources in a cloud environment. the proposed monitoring system was validated by a series of tests that demonstrated it's scalability. All the empirical results were analysed and conclusions were formulated based on the observed properties.
Huge data advent in high-performance computing (HPC) applications such as fluid flow simulations usually hinders the interactive processing and exploration of simulation results. Such an interactive data exploration n...
详细信息
ISBN:
(纸本)9781479984480
Huge data advent in high-performance computing (HPC) applications such as fluid flow simulations usually hinders the interactive processing and exploration of simulation results. Such an interactive data exploration not only allows scientiest to 'play' withtheir data but also to visualise huge (distributed) data sets in both an efficient and easy way. therefore, we propose an HPC data exploration service based on a sliding window concept, that enables researches to access remote data (available on a supercomputer or cluster) during simulation runtime without exceeding any bandwidth limitations between the HPC back-end and the user front-end.
Gradient-based local optimization has been shown to improve results of genetic programming (GP) for symbolic regression. Several state-of-the-art GP implementations use iterative nonlinear least squares (NLS) algorith...
详细信息
ISBN:
(纸本)9781665465458
Gradient-based local optimization has been shown to improve results of genetic programming (GP) for symbolic regression. Several state-of-the-art GP implementations use iterative nonlinear least squares (NLS) algorithms such as the Levenberg-Marquardt algorithm for local optimization. the effectiveness of NLS algorithms depends on appropriate scaling and conditioning of the optimization problem. this has so far been ignored in symbolic regression and GP literature. In this study we use a singular value decomposition of NLS Jacobian matrices to determine the numeric rank and the condition number. We perform experiments with a GP implementation and six different benchmark datasets. Our results show that rank-deficient and ill-conditioned Jacobian matrices occur frequently and for all datasets. the issue is less extreme when restricting GP tree size and when using many non-linear functions in the function set.
Many applications have modules which could benefit greatly from the massive parallel numericcomputing power provided by GPUs. Renderers, signal processing or simulators are only a few such applications. Due to the we...
详细信息
ISBN:
(纸本)9781479930357
Many applications have modules which could benefit greatly from the massive parallel numericcomputing power provided by GPUs. Renderers, signal processing or simulators are only a few such applications. Due to the weaknesses of the GPUs such as stackless execution model or poor capabilities for pointer exchange withthe host, sometimes is not feasible to convert an entire algorithm for GPU, even if it is highly parallel and some of its parts can be greatly accelerated on GPU. In such situations a programmer should have a framework which allows him to split the code flow of a thread in parts and each of these parts will run on the most suitable computing resource, CPU or GPU. For GPU execution, multiple data from host threads will be collected, run on GPU and the results returned to the original threads so they will be able to resume execution on host. In this paper we propose such an algorithm, analyze it and evaluate its practical results.
We investigate the semantics of a language inspired by membrane computing in which computation proceeds in a maximally parallel way, involving multisets of objects distributed into hierarchical structures of regions d...
详细信息
ISBN:
(纸本)9781509057078
We investigate the semantics of a language inspired by membrane computing in which computation proceeds in a maximally parallel way, involving multisets of objects distributed into hierarchical structures of regions delimited by membranes. the language provides primitives for parallel communication of objects across membranes in the form of rules that can be used to express symport/antiport interactions. It also provides a primitive for membrane creation. For the language under investigation we present a denotational semantics designed with metric spaces and continuations.
Grids became a commonplace for the computation of expensive numerical simulations. this work addresses the problem of searching for relevant simulations and for their results in a grid. this is challenging especially ...
详细信息
ISBN:
(纸本)9780769546308
Grids became a commonplace for the computation of expensive numerical simulations. this work addresses the problem of searching for relevant simulations and for their results in a grid. this is challenging especially due to the large number of existing simulations, the small semantic differences between them, and the distributed nature of the grid environment. We propose a solution that addresses simultaneously these three challenges by integrating a latent semantic indexing algorithm a linguistic processing module with a grid application framework. this resulted in a novel prototype, ELSIGExplorer, capable of retrieving relevant scenarios computed with GridSFEA on heterogeneous grids. We evaluated our approach on benchmark datasets from the medical domain and on a set of scenarios for simulating dynamic behavior of biological neural microcircuits in grid.
We describe arithmetic computations in terms of operations on some well known free algebras (S1S, S2S and ordered rooted binary trees) while emphasizing the common structure present in all of them when seen as isomorp...
详细信息
ISBN:
(纸本)9780769549347;9781467350266
We describe arithmetic computations in terms of operations on some well known free algebras (S1S, S2S and ordered rooted binary trees) while emphasizing the common structure present in all of them when seen as isomorphic withthe set of natural numbers. Constructors and deconstructors seen through an initial algebra semantics are generalized to recursively defined functions obeying similar laws. Implementations using GHC's view construct are discussed, based on the free algebra of rooted ordered binary trees.
暂无评论