Besides the inherent challenges of distributedcomputing, gridcomputing presents some new challenges including, how the research professionals, who are not computer expert and normal users who just want to get benefi...
详细信息
ISBN:
(纸本)9781932415575
Besides the inherent challenges of distributedcomputing, gridcomputing presents some new challenges including, how the research professionals, who are not computer expert and normal users who just want to get benefit from grid without taking care of its complexities can be motivated to adopt new and more complex way of working. In this paper we have presented our idea and architecture of 'Users-grid'. Utilizing Agent Technology and with its Service-Oriented architecture it will provide computing power transparently and easily. It gives the feature of true virtualization by enabling existing applications to access grid resources and to shift the extra computational load from users desktop to the other nodes of the grid seamlessly, and bill back the user. It will be helpful to not only research communities and academia but also to business users.
Integer sorting is a subclass of the sorting problem where the elements have integer values and the largest element is polynomially bounded in the number of elements to be sorted. It is useful for applications in whic...
详细信息
ISBN:
(纸本)0818680679
Integer sorting is a subclass of the sorting problem where the elements have integer values and the largest element is polynomially bounded in the number of elements to be sorted. It is useful for applications in which the size of the maximum value of element to be sorted is bounded. In this paper, we present a new distributed radix-sort algorithm for integer sorting. The structure of our algorithm is similar to radix sort except that it typically requires less number of communication phases. We present experimental results for our algorithm on two distributed memory multiprocessors, the Intel Paragon and the Thinking machine CM-5. These results are compared with two other well known practical parallel sorting algorithms based on radix sort and sample sort. The experimental results show that the distributed radix-sort is competitive with the other two algorithms.
The amount of distributed generation and consumption is growing worldwide and consequently also the need to increase the hosting capacity of distribution systems. To achieve this, possible measures include building ne...
详细信息
ISBN:
(纸本)9781538680995
The amount of distributed generation and consumption is growing worldwide and consequently also the need to increase the hosting capacity of distribution systems. To achieve this, possible measures include building new lines, enhancing the grid infrastructure, and providing higher efficiency of the existing grid. Several projects propose to include the control of power generation and Demand-Side-Management, like solar photovoltaic power curtailment and managing flexible demand, as a measure to increase the hosting capacity of the grid. The use of automated demand response through Information and Communications Technologies seems to be a promising solution, which can be rapidly implemented cost-effectively in distribution systems. Automated Demand Response includes a set of activities aimed to reduce or shift electricity use to improve the grid operation. Generators and consumers may be encouraged by the right incentives to activate their flexibilities and consequently increase the hosting capacity of the electricity distribution systems. For this reason, smart grid technologies are expected to play an essential role in the development of future power systems. In this paper, we present the results of different European case studies, which investigated the impact of market-driven automated flexibility management in smart grids. The findings indicate that automated Demand Response can contribute to distribution system stability and increase the renewable energy share, but also requires high computational power and complex coordination architectures.
Simulation has been proved to be a necessary and effective tool for the study and analysis of modern power system. In area interconnected power networks, large-scale real time simulations have become essential. Both d...
详细信息
ISBN:
(纸本)9781424401109
Simulation has been proved to be a necessary and effective tool for the study and analysis of modern power system. In area interconnected power networks, large-scale real time simulations have become essential. Both dynamic simulation (also known as physical simulation) and digital simulation (also called as mathematical simulation) have strong and weak points Based on the complementarities of dynamic simulation and digital simulation, hybrid simulation system can complete accurate and flexible simulations of limited-scale power networks. However, users may face some difficulties when simulation scale exceeding the limitation. gridcomputing can be introduced to resolve the problem. In this paper a seamless-link method of hybrid simulation system based on gridcomputing is presented, which comprises a dynamic simulation system interfaced with the real-time digital simulator (RTDS). At the same time, it can utilize distributedcomputing resource for enlarging the scale of simulation to sufficiently robust to adapt simulation behavior as required.
The success of web services has influenced the way in which grid applications are being written. Web services are increasingly used as a means to realise service-oriented distributedcomputing. grid users often submit...
详细信息
ISBN:
(纸本)9783540683797
The success of web services has influenced the way in which grid applications are being written. Web services are increasingly used as a means to realise service-oriented distributedcomputing. grid users often submit their applications in the form of workflows with certain Quality of Service (QoS) requirements imposed on the workflows. These workflows detail the composition of web services and the level of service required from the grid. This paper addresses workload allocation techniques for grid workflows. We model a web service as an M/M/k queue and obtain a numerical solution for missed deadlines (failures) of grid workflow tasks. The approach is evaluated through an experimental simulation and the results confirm that the proposed workload allocation strategy performs considerably better in terms of satisfying QoS requirements of grid workflows than scheduling algorithms that don't employ such workload allocation techniques.
Generally, smart devices, such as smartphones, smartwatches, or fitness trackers, communicate with each other indirectly, via cloud data centers. Sharing sensor data with a cloud data center as intermediary invokes tr...
详细信息
ISBN:
(纸本)9781538637906
Generally, smart devices, such as smartphones, smartwatches, or fitness trackers, communicate with each other indirectly, via cloud data centers. Sharing sensor data with a cloud data center as intermediary invokes transmission methods with high battery costs, such as 4G LTE or WiFi. By sharing sensor information locally and without intermediaries, we can use other transmission methods with low energy cost, such as Bluetooth or BLE. In this paper, we introduce Sense Low Energy (SenseLE), a decentralized sensing framework which exploits the spatial locality of nearby sensors to save energy in Internet-of-Things (IoT) environments. We demonstrate the usability of SenseLE by building a real-life application for estimating waiting times at queues. Furthermore, we evaluate the performance and resource utilization of our SenseLE Android implementation for different sensing scenarios. Our empirical evaluation shows that by exploiting spatial locality, SenseLE is able to reduce application response times (latency) by up to 74% and energy consumption by up to 56%.
There is a growing interest in training deep neural networks (DNNs) in a GPU cloud environment. This is typically achieved by running parallel training workers on multiple GPUs across computing nodes. Under such a set...
详细信息
ISBN:
(数字)9781665471770
ISBN:
(纸本)9781665471770
There is a growing interest in training deep neural networks (DNNs) in a GPU cloud environment. This is typically achieved by running parallel training workers on multiple GPUs across computing nodes. Under such a setup, the communication overhead is often responsible for long training time and poor scalability. This paper presents AIACC-Training, a unified communication framework designed for the distributed training of DNNs in a GPU cloud environment. AIACC-Training permits a training worker to participate in multiple gradient communication operations simultaneously to improve network bandwidth utilization and reduce communication latency. It employs auto-tuning techniques to dynamically determine the right communication parameters based on the input DNN workloads and the underlying network infrastructure. AIACC-Training has been deployed to production at Alibaba GPU Cloud with 3000+ GPUs executing AIACC-Training optimized code at any time. Experiments performed on representative DNN workloads show that AIACC-Training outperforms existing solutions, improving the training throughput and scalability by a large margin.
As the large volume of resources involved and the power of computational grids increased, there is a corresponding and urgent need for employ the grid technologies into problem solving environment (PSE) domain in orde...
详细信息
A parallel programming paradigm dictates the way an application is to be expressed. It also restricts the algorithms that may be used in the application. Unfortunately, runtime systems for parallelcomputing often imp...
详细信息
ISBN:
(纸本)0769517455
A parallel programming paradigm dictates the way an application is to be expressed. It also restricts the algorithms that may be used in the application. Unfortunately, runtime systems for parallelcomputing often impose a particular programming paradigm. For a wider choice of algorithms, it is desirable to support more than one paradigm. In this paper we propose our work in SilkRoad II, a variant of the Cilk runtime system for cluster computing. What is unique about SilkRoad II is its memory model which supports multiple paradigms with the underlying software distributed shared memory. The RC_dag memory consistency model of SilkRoad H is introduced. Our experimental results show that the stronger RC_dag can achieve performance comparable to LC of Cilk while supporting a bigger set of paradigms with rather good performance.
The main objective of this work is to quantitatively study the extraction problem of distributed parameters of unparallel transmission lines, especially for one specific type of unparallel transmission line model. The...
详细信息
ISBN:
(纸本)9781479952748
The main objective of this work is to quantitatively study the extraction problem of distributed parameters of unparallel transmission lines, especially for one specific type of unparallel transmission line model. The infinitesimal transmission lines are equivalent to the parallel transmission lines, and the estimation method of distributed parameters is presented based on the original physical definition of distributed parameters. Numerical experiments show that distributed parameters are related to axial variable and the angle of wires, it represents the nonuniform characteristic of parameters;when the angle of the wires is close to zero, the distributed parameters are the same as the parameters of parallel transmission lines, which validates the effectiveness of the method.
暂无评论