Though there is increased activity in the implementation of asymptotically fast polynomial arithmetic, little is reported on the details of such effort. In this paper, we discuss how we achieve high performance in imp...
详细信息
ISBN:
(纸本)1595932763
Though there is increased activity in the implementation of asymptotically fast polynomial arithmetic, little is reported on the details of such effort. In this paper, we discuss how we achieve high performance in implementing some well-studied fast algorithms for polynomial arithmetic in two high-level programming environments, AXIOM and ALDOR. Two approaches are investigated. With ALDOR we rely only on high-level generic code, whereas with AXIOM we endeavor to mix high-level, middle-level and low-level specialized code. We show that our implementations are satisfactory compared with other known computer algebra systems or libraries such as MAGMA v2.11-2 and NTL v5.4. Copyright 2006 ACM.
Modern, resource-intensive enterprise and scientific applications are increasingly architected using parallel and distributed components deployed on grid platforms. Grid environments are typically heterogeneous and su...
详细信息
ISBN:
(纸本)0769525857
Modern, resource-intensive enterprise and scientific applications are increasingly architected using parallel and distributed components deployed on grid platforms. Grid environments are typically heterogeneous and subject to large load variations. Such characteristics require a portable programming infrastructure and comprehensive support for resource management as well as dynamic adaptation to varying resource availability. Resource-aware applications pro-actively monitor and dynamically adapt their behavior to accommodate to changing resource consumption constraints. Strategies such as internal load-balancing and job resubmission can correct sub-optimal scheduling decisions and thus enhance performance. As utilizing the potential of resource awareness requires close interaction with resource management subsystems, implementations often rely on low-level, platform-specific features, which impairs portability. What is needed is uncompromised portability of all aspects of resource management in grid systems. We demonstrate that the Java platform, equipped with the APIs for distributed resource management, can be used for building a portable grid infrastructure. In particular, we describe the design and implementation of a cluster management system capable of controlled execution of parallel, resource aware, Java applications. We empirically show that our solution allows for efficient, fine-grained resource management, sufficient for tapping the performance benefits promised by the resource-conscious applications.
Speculation techniques in the design of superscalar processors and multiprocessors allow processors to divide a program's execution automatically and dynamically into several execution threads. They also increase ...
详细信息
ISBN:
(纸本)9781622763535
Speculation techniques in the design of superscalar processors and multiprocessors allow processors to divide a program's execution automatically and dynamically into several execution threads. They also increase the degree of parallelism of a program thanks to speculation on data and control dependences. However, cluster based systems allow parallel programs to be executed based on packages such as PVM and MPI. These packages need explicit parallel programming and do not have provisions for automatic parallelisation or speculation. These two previous premises together with Moore's law (processing power doubles every 18 months) and Gilder's law (bandwidth triples every 12 months) make the idea of transporting speculation techniques to a distributed environment of a low cost workstation possible. To do this, we should be able to divide programs and build a graph of instruction blocks, where the input and output data of each block is known. Exploiting existing parallelism and using speculation techniques should provide better performance than sequential execution on a single workstation. In this paper we analyse the viability of the described proposal using simulations.
In this paper we study multi issue alternating-offers bargaining in a perfect information finite horizon setting, we determine the pertinent subgame perfect equilibrium, and we provide an algorithm to compute it. The ...
详细信息
In this paper we study multi issue alternating-offers bargaining in a perfect information finite horizon setting, we determine the pertinent subgame perfect equilibrium, and we provide an algorithm to compute it. The equilibrium is determined by making a novel use of backward induction together with convex programming techniques in multi issue settings. We show that the agents reach an agreement immediately and that such an agreement is Pareto efficient. Furthermore, we prove that, when the multi issue utility functions are linear, the problem of computing the equilibrium is tractable and the related complexity is polynomial with the number of issues and linear with the deadline of bargaining.
Image and video codecs are prevalent in multimedia devices, ranging from embedded systems, to desktop computers, to high-end servers such as HDTV editing consoles. It is not uncommon however that developers create and...
详细信息
We present a real-time reverse radiosity method for compensating indirect scattering effects that occur with immersive and semi-immersive projection displays. It computes a numerical solution directly on the GPU and i...
详细信息
ISBN:
(纸本)1424402239
We present a real-time reverse radiosity method for compensating indirect scattering effects that occur with immersive and semi-immersive projection displays. It computes a numerical solution directly on the GPU and is implemented with pixel shading and multi-pass rendering which together realizes a Jacobi solver for sparse matrix linear equation systems. Our method is validated and evaluated based on a stereoscopic two-sided wall display. The images appear more brilliant and uniform when compensating the scattering contribution.
The Java language and platform have been considered by many as natural candidate for creating grid systems. The platform-independent runtime environment, safe and high-level language and its built-in support for netwo...
详细信息
Writing a perfectly correct code is a challenging and a nearly impossible task. In this work we suggest the recovery oriented programming paradigm in order to cope with eventual Byzantine programs. The program specifi...
详细信息
ISBN:
(纸本)3540490183
Writing a perfectly correct code is a challenging and a nearly impossible task. In this work we suggest the recovery oriented programming paradigm in order to cope with eventual Byzantine programs. The program specification composer enforces the program specifications (both the safety and the liveness properties) in run time using predicates over input and output variables. The component programmer will use these variables in the program implementation. We suggest using the "sand-box" approach in which every instruction of the program that changes a specification variable, is executed first with temporary variables and that is in order to avoid execution of an instruction that violates the specifications. In addition, external monitoring is used for coping with transient faults and for ensuring convergence to a legal state. The implementation of these ideas includes the definition of new instructions in the programming language with the purpose of allowing addition of predicates and recovery actions. We suggest a design for a tool that extends the Java programming language. In addition to that, we provide a correctness proof scheme for proving that the code combined with the predicates and the recovery actions is self-stabilizing and, under the restartability assumption, eventually fulfills its specifications.
Agent oriented software development is generally motivated by the need of open and adaptive systems development that continuously change and evolve to meet new requirements. In this paper, we propose an agent-based op...
详细信息
ISBN:
(纸本)0769525466
Agent oriented software development is generally motivated by the need of open and adaptive systems development that continuously change and evolve to meet new requirements. In this paper, we propose an agent-based open and adaptive system development process. The proposed methodology is based on model-based techniques and provides a specific model for the type of information to he gathered and uses this model to drive the user-oriented requirement analysis, specification and system-oriented analysis process. The proposed methodology also reduces the complexity of the software system and provides better quality with better user-satisfaction. The Gaia and ROADMAP models have been used as a fundamental ingredient to the new proposed agent-based modeling method In the proposed agent-based system, the focus is on a clear separtion between the requirement gathering and analysis phases. The scope of the discussion covers requirements gathering and user-oriented analysis phases only and does not touch upon the system-analysis and design phases.
It is well known that in any ab initio molecular orbital (MO) calculation, the major task involves the computation of molecular integrals, among which the computation of three-center nuclear attraction and Coulomb int...
详细信息
It is well known that in any ab initio molecular orbital (MO) calculation, the major task involves the computation of molecular integrals, among which the computation of three-center nuclear attraction and Coulomb integrals is the most frequently encountered. As the molecular system becomes larger, computation of these integrals becomes one of the most laborious and time-consuming steps in molecular systems calculation. Improvement of the computational methods of molecular integrals would be indispensable to further development in computational studies of large molecular systems. To develop fast and accurate algorithms for the numerical evaluation of these integrals over B functions, we used nonlinear transformations for improving convergence of highly oscillatory integrals. These methods form the basis of new methods for solving various problems that were unsolvable otherwise and have many applications as well. To apply these nonlinear transformations, the integrands should satisfy linear differential equations with coefficients having asymptotic power series in the sense of Poincare, which in their turn should satisfy some limit conditions. These differential equations are very difficult to obtain explicitly. In the case of molecular integrals, we used a symbolic programming language (MAPLE) to demonstrate that all the conditions required to apply these nonlinear transformation methods are satisfied. Differential equations are obtained explicitly, allowing us to demonstrate that the limit conditions are also satisfied. (c) 2005 Wiley Periodicals, Inc.
暂无评论