Desktop computing remains indispensable in scientific exploration, largely because it provides people with devices for human interaction and environments for interactive job execution. However, with today's rapidl...
详细信息
Desktop computing remains indispensable in scientific exploration, largely because it provides people with devices for human interaction and environments for interactive job execution. However, with today's rapidly growing data volume and task complexity, it is increasingly hard for individual workstations to meet the demands of interactive scientific data processing. The increasing cost of such interactive processing is hindering the productivity of end-to-end scientific computing workflows. While existing distributedcomputing systems allow people to aggregate desktop workstation resources for parallelcomputing, the burden of explicit parallel programming and parallel job execution often prohibits scientists to take advantage of such platforms. In this paper, we discuss the need for transparent desktop parallelcomputing in scientific data processing. As an initial step toward this goal, we present our on-going work on the automatic parallelization of the scripting language R, a popular tool for statistical computing. Our preliminary results suggest that a reasonable speedup can be achieved on real-world sequential R programs without requiring any code modification.
The aim of this paper is to present two new portable and high performance implementations of routines that can be used for piecewise cubic interpolation. The first one (sequential) is based on LAPACK routines, while t...
详细信息
The aim of this paper is to present two new portable and high performance implementations of routines that can be used for piecewise cubic interpolation. The first one (sequential) is based on LAPACK routines, while the next, based on ScaLAPACK is designed for distributed memory parallel computers and clusters. The results of experiments performed on a cluster of twenty Itanium 2 processors and on Cray XI are also presented and shortly discussed
This article describes a new re-configurable architecture design suitable for high performance signal processing applications. The main benefits beside re-configurability are high computation performance, robustness a...
详细信息
Computer vision and image processing has always been an active research domain that requires an enormous computational effort. In this paper we present the architecture of a modular object retrieval system. It is base...
详细信息
Computer vision and image processing has always been an active research domain that requires an enormous computational effort. In this paper we present the architecture of a modular object retrieval system. It is based on a dataflow concept which allows flexible adaptations to different tasks. This concept facilitates parallel processing as well as distributedcomputing. We also present a dynamic load balancing service for heterogeneous environments that has been integrated to improve system performance. First experiments show that the developed balancer performs better than standard balancing techniques in this environment.
Techniques based on metaheuristics and nature-inspired paradigms can provide efficient solutions to a wide variety of problems. Moreover, parallel and distributed metaheuristics can be used to provide more powerful pr...
详细信息
Techniques based on metaheuristics and nature-inspired paradigms can provide efficient solutions to a wide variety of problems. Moreover, parallel and distributed metaheuristics can be used to provide more powerful problem solving environments in a variety of fields, ranging, for example, from finance to bio- and health-informatics. This workshop seeks to provide an opportunity for researchers to explore the connection between metaheuristics and the development of solutions to problems that arise in operations research, parallelcomputing, telecommunications, and many others.
This paper presents a novel algorithm to detect null pointer dereference errors. The algorithm utilizes both of the must and may alias information in a compact way to improve the precision of the detection. Using may ...
详细信息
ISBN:
(纸本)9783540884781
This paper presents a novel algorithm to detect null pointer dereference errors. The algorithm utilizes both of the must and may alias information in a compact way to improve the precision of the detection. Using may alias information obtained by a fast flow- and context- insensitive analysis algorithm, we compute the must alias generated by the assignment statements and the must alias information is also used to improve the precision of the may alias. We can strong update more expressions using the must alias information, which will reduce the false positives of the detection for null pointer dereference. We have implemented our algorithm in the SUIF2 compiler infrastructure and the experiments results are as expected.
In the wide-area Grid environment which consists of a huge amount of stateful Grid services, the service discovery is a key and challenging issue. The user's requirements on Grid services not only include the func...
详细信息
ISBN:
(纸本)3540241280
In the wide-area Grid environment which consists of a huge amount of stateful Grid services, the service discovery is a key and challenging issue. The user's requirements on Grid services not only include the function-related but also include the QoS or service state related requirements. Based on the analysis for the requirements of service discovery, this paper divides the service matching process into 2 stages: service type matching and instance matching, and proposes a Grid Service Discovery Model Based on 2-Stage Matching which can enable more effective service discovery. In the model VO is utilized as the managerial unit for grid services and a two-level publication architecture is adopted. The initial simulation results show that the model can effectively aggregate the service information and avoid the workload caused by frequent dynamic updating.
In this poster, the energy and carbon management problem in Data Center Microgrid (DCMG) is modeled as a Decentralized Partially Observable Markov Decision Process, and the Multi-Agent Deep Deterministic Policy Gradie...
详细信息
It is our pleasure to welcome all speakers, authors and participants to this fifth edition of the international Workshop on System Management Techniques, Processes, and Services (SMTPS), organized as a 2009 IPDPS (IEE...
详细信息
It is our pleasure to welcome all speakers, authors and participants to this fifth edition of the international Workshop on System Management Techniques, Processes, and Services (SMTPS), organized as a 2009 IPDPS (IEEE internationalparallel & distributed Processing symposium) Workshop.
暂无评论