the Global Arrays toolkit is a library that allows programmers to write parallel programs that use large arrays distributed across processing nodes. the toolkit provides advantage of one-sided communication operations...
详细信息
ISBN:
(纸本)9781424421015
the Global Arrays toolkit is a library that allows programmers to write parallel programs that use large arrays distributed across processing nodes. the toolkit provides advantage of one-sided communication operations through the Aggregate Remote memory Copy Interface (ARMCI). In this study, we investigate the performance of a parallel application implemented with (he toolkit on cluster and Grid computing environments as well as comparison of performance of same application implemented with MPI library. In cluster computing environment, the performance of the evaluating application implemented withthe toolkit outperforms MPICH implementation when message size is large enough. However, for Grid computing environment with lower bandwidth and high latency conditions, the performance of the evaluating application implemented withthe toolkit is significantly dropped compared to that of the MPICH-G2 implementation.
this paper describes a highly flexible component architecture, primarily designed for automotive control systems, that supports distributed dynamically-configurable context-aware behaviour. the architecture enforces a...
详细信息
Traditional outlier detection algorithms are designed to apply to centralized databases, not distributed databases. We proposed a novel outlier detection algorithm for distributed databases. Given data assigned to dif...
详细信息
ISBN:
(纸本)9780769533056
Traditional outlier detection algorithms are designed to apply to centralized databases, not distributed databases. We proposed a novel outlier detection algorithm for distributed databases. Given data assigned to different network nodes of a network platform, where each node has its own memory and hard disc, and the communication between nodes driven by message, the populated data would be non-overlapping. the working way of the network system is a manager-worker mode, that is, that a node as manager is responsible for assigning tasks to worker and querying the results from worker nodes. the algorithm first detected local outliers based on distance on all nodes, and then identified local outliers collected in the central node where a globally screening operation on all local outliers was implemented to achieve really global outliers. To scale the algorithm to massive data and reduce its computing complexity, a data filtering technology was further presented. Experimental results demonstrated that the algorithm effectively and efficiently handled on real and artificial data.
the global arrays toolkit is a library that allows programmers to write parallel programs that use large arrays distributed across processing nodes. the toolkit provides advantage of one-sided communication operations...
详细信息
the global arrays toolkit is a library that allows programmers to write parallel programs that use large arrays distributed across processing nodes. the toolkit provides advantage of one-sided communication operations through the aggregate remote memory copy interface (ARMCI). In this study, we investigate the performance of a parallel application implemented withthe toolkit on cluster and grid computing environments as well as comparison of performance of same application implemented with MPI library. In cluster computing environment, the performance of the evaluating application implemented withthe toolkit outperforms MPICH implementation when message size is large enough. However, for grid computing environment with lower bandwidth and high latency conditions, the performance of the evaluating application implemented withthe toolkit is significantly dropped compared to that of the MPICH-G2 implementation.
Efficient data management techniques are needed in wireless sensor networks (WSNs) to counteract issues related to limited resources, e.g. energy, memory, bandwidth, as well as limited connectivity. Self-organizing an...
详细信息
Efficient data management techniques are needed in wireless sensor networks (WSNs) to counteract issues related to limited resources, e.g. energy, memory, bandwidth, as well as limited connectivity. Self-organizing and cooperative algorithms are thought to be the optimal solution to overcome these limitations. On an abstract level, structured peer-to-peer protocols provide O(1) complexity for storing and retrieving data in the network. However, they rely on underlayer routing techniques. In this paper, we present the virtual cord protocol (VCP), a virtual relative position based efficient routing protocol that also provides means for data management, e.g. insert, get, and delete, as known from typical distributed hash table (DHT) services. the key contributions of this protocol are independence of real location information by relying on relative positions of neighboring nodes, short virtual paths because successors and predecessors are in their vicinity, and high scalability because only information about direct neighbors is needed for routing. Furthermore, VCP inherently prevents dead-ends and it is easy to be implemented.
the proceedings contain 20 papers. the topics discussed include: rational models of cognitive control;fault-tolerance in biochemical systems;optical computing and computational complexity;handling Markov chains with m...
详细信息
ISBN:
(纸本)3540385932
the proceedings contain 20 papers. the topics discussed include: rational models of cognitive control;fault-tolerance in biochemical systems;optical computing and computational complexity;handling Markov chains with membrane computing;approximation classes for real number optimization problems;physical systems as constructive logics;chemical information processing devices constructed using a nonlinear medium with controlled excitability;flexible versus rigid tile assembly;the number of orbits of periodic box-ball systems;optimizing potential information transfer with self-referential memory;on the power of bio-turing machines;and ergodic dynamics for large-scale distributed robot systems.
Withthe development of grid techniques and the growing complexity of grid applications, reasoning the temporal properties of grid application to ensure its reliability is becoming more and more critical. In this work...
详细信息
ISBN:
(纸本)9780769526942
Withthe development of grid techniques and the growing complexity of grid applications, reasoning the temporal properties of grid application to ensure its reliability is becoming more and more critical. In this work, two decomposition approaches are proposed to improve the performance of the temporal reasoning of complex grid applications. the proposed approaches are implemented in our GridPiAnalyzer for equipment grid. Results show that our approach can reduce both CPU time and memory cost compared to using traditional formal verification algorithm alone due to the exponential reduction of system state space.
the paper describes parallel 3D DEM simulation of compacting of heterogeneous material. Static domain decomposition and message passing inter-processor communication have been implemented in the DEM code. A novel algo...
详细信息
A method for computing orthogonal URV/ULV decompositions of block tridiagonal (or banded) matrices is presented. the method discussed transforms the matrix into structured triangular form and has several attractive pr...
详细信息
Most parallel computations require the exchange of data between processing elements. One of important basic communication operations is all-reduce, a variation of the reduction operation. this paper presents an all-re...
详细信息
ISBN:
(数字)9783540321118
ISBN:
(纸本)3540260323
Most parallel computations require the exchange of data between processing elements. One of important basic communication operations is all-reduce, a variation of the reduction operation. this paper presents an all-reduce communication operation scheme using all-to-all broadcast communication pattern. All-to-all broadcast is the operation in which each processor sends its message to all other processors, and receives messages from all other processors in the system. In this paper, we develop an efficient all-reduce operation scheme in a star network topology withthe single-port communication capability. Communication time is compared against known broadcasting schemes to verify the efficiency of the suggested scheme.
暂无评论