An evaluation for a parallel Particle-In-Cell code leveraging heterogeneous hardware is presented. Two parallelization strategies are investigated on Intel R © Xeon Phi TM coprocessors. Hybrid parallelization is implemented to support optional workload offloading to coprocessors. A performance model is applied to load balance heterogeneous setups. Performance measurements of a benchmark show th...

The combination of inter-node and intra-node domaindecomposition strategies for the development of memorysaving parallel Particle-in-cell simulation codes, targeted to hierarchical distributed-shared memory architectures, is discussed in this paper, along with its MPI+OpenMP implementation. Particular emphasis is given to the devised dynamic workload balancing technique.

In recent years, several frameworks have been developed for processing very large quantities of data on large clusters of commodity PCs. These frameworks have focused on fault-tolerance and scalability. However, when using heterogeneous environments these systems do not offer optimal workload balancing. In this paper we present Jumbo, a distributed computation platform designed to explore possi...

MapReduce has become a popular model for large-scale data processing in recent years. Many works on MapReduce scheduling (e.g., load balancing and deadline-aware scheduling) have emphasized the importance of predicting workload received by individual reducers. However, because the input characteristics and user-specified map function of a given job are unknown to the MapReduce framework before ...

OF THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science Electrical Engineering The University of New Mexico Albuquerque, New Mexico June, 2004 Implementation of Load Balancing Policies in Distributed Systems by Jean Ghanem B.E., American University of Beirut, 2002 M.S., Electrical Engineering, University of New Mexico, 2004 Abstract Load balancing is t...

Although the simple assembly line balancing problem (SALBP) is the topic of many studies, typically they either consider minimizing the number of stations for a given cycle time (called type one), or minimizing the cycle time for a given number of stations (called type two). Rarely, type E of the problem has been considered. In the type E, cycle time and number of stations are both decision var...

As Sensor area network is a vast network with thousands of nodes, the workload increases up to a great exten t. Wireless sensor networks are increasingly gaining importance in various time critical applications. As a sensor node has fix power to work on, when the load exceeds its threshold value, it starts behaving abnormally and starts packet as well as energ y loss while the corner nodes are ...

The feature of load sharing or load balancing involves migration of running processes from highly loaded workstations of a network to the lightly-loaded or idle workstations of the network. This paper describes load balancing techniques to share the workload of the workstations belonging to a particular network to gain better performance from the overall network. The mechanisms of load informat...

This paper addresses the coverage and workloadbalancing requirements of worker recruiting in spatial crowdsourcing. That is, the recruited workers should be able to visit all the crowdsourcing locations to satisfy a certain quality, e.g., traffic monitoring or climate forecast. In addition, each crowdsourcing operation has a cost, e.g., data traffic or energy consumption, and each crowdsourcing...

Grid computing systems are distributed systems that involve coordinate and involvement of heterogeneous resources with various characteristics where user jobs can be executed on either local or remote computer. These heterogeneous computing resources are used to run highly complex programs that require very high processing power and huge volume of input data. Recently the biggest issue in distr...