Results 1  10
of
148,368
Improving PerNode Efficiency in the Datacenter with New OS Abstractions
"... We believe datacenters can benefit from more focus on pernode efficiency, performance, and predictability, versus the more common focus so far on scalability to a large number of nodes. Improving pernode efficiency decreases costs and fault recovery because fewer nodes are required for the same am ..."
Abstract

Cited by 10 (0 self)
 Add to MetaCart
We believe datacenters can benefit from more focus on pernode efficiency, performance, and predictability, versus the more common focus so far on scalability to a large number of nodes. Improving pernode efficiency decreases costs and fault recovery because fewer nodes are required for the same
A Scalable Model for Channel Access Protocols in Multihop Ad Hoc Networks
, 2004
"... A new modeling framework is introduced for the analytical study of medium access control (MAC) protocols operating in multihop ad hoc networks. The model takes into account the e#ect of physicallayer parameters on the success of transmissions, the MAC protocol on the likelihood that nodes can acces ..."
Abstract

Cited by 83 (5 self)
 Add to MetaCart
access the channnel, and the connectivity of nodes in the network. A key feature of the model is that nodes can be modeled individually, i.e., it allows a pernode setup of many layerspecific parameters. Moreover, no spatial probability distribution or a particular arrangement of nodes is assumed
Optimal perNode Rate Allocation to provide perFlow EndtoEnd Delay Guarantees in a Network of Routers supporting Guaranteed Service Class
, 2001
"... Abstract—In this paper, we investigate the problem of providing worstcase endtoend delay guarantee to a token bucket constrained flow traversing a series of N packet schedulers. We consider a network of routers that support the Guaranteed Service class of the IETF Integrated Services (IntServ) Wo ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
Abstract—In this paper, we investigate the problem of providing worstcase endtoend delay guarantee to a token bucket constrained flow traversing a series of N packet schedulers. We consider a network of routers that support the Guaranteed Service class of the IETF Integrated Services (IntServ) Working Group; this service class is proposed to provide Quality of Service (QoS) in the Internet. Under this framework, a worstcase endtoend delay bound to a flow is provided by allocating a rate at each network element on the path of the flow. We associate a cost with allocating a rate at each link. The cost is assumed to be a convex and nondecreasing function of rate. We investigate the problem of obtaining an optimal rate allocation for a flow that minimizes the total cost subject to the delay requirement and the available link capacity constraints. Allocating an identical rate at each link on the path of a flow is a widely used approach under the Guaranteed Service framework. We investigate the optimality of this approach and show that under certain conditions, it need not be optimal. Moreover, we investigate the optimal solution to the total cost minimization problem and give scenarios in which we can explicitly obtain the optimal solution. Based on these results, we present an algorithm for optimal rate allocation that is based on multiple rates. However, with blocking probability as the performance criterion, we find through simulations, that the optimal rate allocation algorithm is only marginally better for connections with longer path lengths at the expense of those with shorter path lengths. We also observe that the performances of both the algorithms are very close and so, in practice, the simpler identical rate algorithm may be sufficient. I.
Discovery of Inference Rules for Question Answering
 Natural Language Engineering
, 2001
"... One of the main challenges in questionanswering is the potential mismatch between the expressions in questions and the expressions in texts. While humans appear to use inference rules such as “X writes Y ” implies “X is the author of Y ” in answering questions, such rules are generally unavailable ..."
Abstract

Cited by 307 (7 self)
 Add to MetaCart
One of the main challenges in questionanswering is the potential mismatch between the expressions in questions and the expressions in texts. While humans appear to use inference rules such as “X writes Y ” implies “X is the author of Y ” in answering questions, such rules are generally unavailable to questionanswering systems due to the inherent difficulty in constructing them. In this paper, we present an unsupervised algorithm for discovering inference rules from text. Our algorithm is based on an extended version of Harris ’ Distributional Hypothesis, which states that words that occurred in the same contexts tend to be similar. Instead of using this hypothesis on words, we apply it to paths in the dependency trees of a parsed corpus. Essentially, if two paths tend to link the same set of words, we hypothesize that their meanings are similar. We use examples to show that our system discovers many inference rules easily missed by humans. 1
GXP: An Interactive Shell for the Grid Environment
"... Abstract — We describe GXP, a shell for distributed multicluster environments. With GXP, users can quickly submit a command to many nodes simultaneously (approximately 600 milliseconds on over 300 nodes spread across five localarea networks). It therefore brings an interactive and instantaneous res ..."
Abstract
 Add to MetaCart
command and all parallel commands), and (3) a flexible and efficient method to interactively select a subset of nodes to execute subsequent commands on. It is very easy to start using GXP, because it is designed not to require cumbersome pernode setup and installation and to depend only on a very small
Efficient Setup . . .
, 2007
"... Solving partial differential equations (PDEs) using analytical techniques is intractable for all but the simplest problems. Many computational approaches to approximate solutions to PDEs yield large systems of linear equations. Algorithms known as linear solvers then compute an approximate solution ..."
Abstract
 Add to MetaCart
Solving partial differential equations (PDEs) using analytical techniques is intractable for all but the simplest problems. Many computational approaches to approximate solutions to PDEs yield large systems of linear equations. Algorithms known as linear solvers then compute an approximate solution to the linear system. Multigrid methods are one class of linear solver and find an approximate solution to a linear system through two complementary processes: relaxation and coarsegrid correction. Relaxation cheaply annihilates portions of error from the approximate solution, while coarsegrid correction constructs a lower dimensional problem to remove error remaining after relaxation. In algebraic multigrid (AMG), the lower dimensional space is constructed by coarsegrid selection algorithms. In this thesis, an introduction and study of independent setbased parallel coarsegrid selection algorithms is presented in detail, following a review of algebraic multigrid. The behavior of the ClearyLubyJonesPlassmann (CLJP) algorithm is analyzed and modifications to the initialization phase of CLJP are recommended, resulting in the CLJP in Color (CLJPc) algorithm, which achieves large performance gains over CLJP for problems on uniform grids. CLJPc is then extended to the Parallel Modified Independent Set (PMIS) coarsegrid selection algorithm producing
Results 1  10
of
148,368